topo.transferfn Package

---

transferfn Package

A family of function objects for transforming a matrix generated from some other function.

Output functions are useful for neuron activation functions, normalization of matrices, etc. They come in two varieties: OutputFunction, and CFPOutputFunction. An OutputFunction (e.g. PiecewiseLinear) applies to one matrix of any type, such as an activity matrix or a set of weights. To apply an OutputFunction to all of the weight matrices in an entire CFProjection, an OutputFunction can be plugged in to CFPOF_Plugin. CFPOF_Plugin is one example of a CFPOutputFunction, which is a function that works with the entire Projection at once.

Any new output functions added to this directory will automatically become available for any model.

class topo.transferfn.TransferFnWithState(**params)

Bases: imagen.transferfn.TransferFn

Abstract base class for TransferFns that need to maintain a self.plastic parameter.

These TransferFns typically maintain some form of internal history or other state from previous calls, which can be disabled by override_plasticity_state().

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82db8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82d60>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e68>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82e10>

set_param = <functools.partial object at 0x2b07b1f82fc8>

state_pop()

Restore the state saved by the most recent state_push call.

state_push()

Save the current state onto a stack, to be restored with state_pop.

Subclasses must implement state_push and state_pop to store state across invocations. The behaviour should be such that after state_pop, the state is restored to what it was at the time when state_push was called.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.TransferFnWithRandomState(**params)

Bases: imagen.transferfn.TransferFnWithState, numbergen.TimeAwareRandomState

Abstract base class for TransferFns that use a random state. Inherits time-dependent control of the random state from numbergen.TimeAwareRandomState. Consult the help of TimeAwareRandomState for more information.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Parameter random_generator (allow_None=False, constant=False, default=<mtrand.RandomState object at 0x2b07ab473c50>, instantiate=False, pickle_default_value=True, precedence=-1, readonly=False)

Using Numpy’s RandomState instead of random.Random as the former can generate random arrays and more random distributions. See RandomState’s help for more information.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Boolean time_dependent (allow_None=False, bounds=(0, 1), constant=False, default=False, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether the given time_fn should be used to constrain the results generated.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Callable time_fn (allow_None=False, constant=False, default=<Time Time00001>, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Callable used to specify the time that determines the state and return value of the object, if time_dependent=True.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82db8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82ba8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82c58>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82fc8>

set_param = <functools.partial object at 0x2b07b1f82f70>

state_pop()

Retrieve the previous random number generator from the stack.

state_push()

Save the current random number generator (onto the stack), replacing it with a copy.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.HalfRectifyAndPower(**params)

Bases: imagen.transferfn.TransferFn

Transfer function that applies a half-wave rectification (i.e., clips at zero), and then raises the result to the e-th power (where the exponent e can be selected arbitrarily).

param Number e (allow_None=False, bounds=None, constant=False, default=2.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The exponent to which the thresholded value is raised.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number t (allow_None=False, bounds=None, constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The threshold level subtracted from x.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82c58>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82f70>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e68>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82b50>

set_param = <functools.partial object at 0x2b07b1f82d08>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.PoissonSample(**params)

Bases: topo.transferfn.TransferFnWithRandomState

Simulate Poisson-distributed activity with specified mean values.

This transfer function interprets each matrix value as the (potentially scaled) rate of a Poisson process and replaces it with a sample from the appropriate Poisson distribution.

To allow the matrix to contain values in a suitable range (such as [0.0,1.0]), the input matrix is scaled by the parameter in_scale, and the baseline_rate is added before sampling. After sampling, the output value is then scaled by out_scale. The function thus performs this transformation:

x <- P(in_scale * x + baseline_rate) * out_scale

where x is a matrix value and P(r) samples from a Poisson distribution with rate r.

param Number baseline_rate (allow_None=False, bounds=None, constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Constant to add to the input after scaling, resulting in a baseline Poisson process rate.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Parameter random_generator (allow_None=False, constant=False, default=<mtrand.RandomState object at 0x2b07ab473c50>, instantiate=False, pickle_default_value=True, precedence=-1, readonly=False)

Using Numpy’s RandomState instead of random.Random as the former can generate random arrays and more random distributions. See RandomState’s help for more information.

param Number in_scale (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Amount by which to scale the input.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Boolean time_dependent (allow_None=False, bounds=(0, 1), constant=False, default=False, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether the given time_fn should be used to constrain the results generated.

param Number out_scale (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Amount by which to scale the output (e.g. 1.0/in_scale).

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Callable time_fn (allow_None=False, constant=False, default=<Time Time00001>, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Callable used to specify the time that determines the state and return value of the object, if time_dependent=True.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82c00>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82db8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82cb0>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82ba8>

set_param = <functools.partial object at 0x2b07b1f82ec0>

state_pop()

Retrieve the previous random number generator from the stack.

state_push()

Save the current random number generator (onto the stack), replacing it with a copy.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.ScalingTF(**params)

Bases: imagen.transferfn.TransferFnWithState

Scales input activity based on the current average activity (x_avg).

The scaling is calculated to bring x_avg for each unit closer to a specified target average. Calculates a scaling factor that is greater than 1 if x_avg is less than the target and less than 1 if x_avg is greater than the target, and multiplies the input activity by this scaling factor.

The plastic parameter allows the updating of the average values to be disabled temporarily, e.g. while presenting test patterns.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number target (allow_None=False, bounds=None, constant=False, default=0.01, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Target average activity for each unit.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Number step (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

How often to calculate the average activity and scaling factor.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number smoothing (allow_None=False, bounds=None, constant=False, default=0.9997, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Determines the degree of weighting of previous activity vs. current activity when calculating the average.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82c58>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82ec0>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e10>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82fc8>

set_param = <functools.partial object at 0x2b07b1f82f70>

state_pop()

Restore the state saved by the most recent state_push call.

state_push()

Save the current state onto a stack, to be restored with state_pop.

Subclasses must implement state_push and state_pop to store state across invocations. The behaviour should be such that after state_pop, the state is restored to what it was at the time when state_push was called.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.PiecewiseLinear(**params)

Bases: imagen.transferfn.TransferFn

Piecewise-linear TransferFn with lower and upper thresholds.

Values below the lower_threshold are set to zero, those above the upper threshold are set to 1.0, and those in between are scaled linearly.

param Number lower_bound (allow_None=False, bounds=None, constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number upper_bound (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82f70>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c58>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e68>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82db8>

set_param = <functools.partial object at 0x2b07b1f82f18>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.TransferFn(**params)

Bases: param.parameterized.Parameterized

Function object to modify a matrix in place, e.g. for normalization.

Used for transforming an array of intermediate results into a final version, by cropping it, normalizing it, squaring it, etc.

Objects in this class must support being called as a function with one matrix argument, and are expected to change that matrix in place.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82ec0>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82f70>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e10>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82d08>

set_param = <functools.partial object at 0x2b07b1f82b50>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.IdentityTF(**params)

Bases: imagen.transferfn.TransferFn

Identity function, returning its argument as-is.

For speed, calling this function object is sometimes optimized away entirely. To make this feasible, it is not allowable to derive other classes from this object, modify it to have different behavior, add side effects, or anything of that nature.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e68>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c58>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82ec0>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82ba8>

set_param = <functools.partial object at 0x2b07b1f82db8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.Hysteresis(**params)

Bases: imagen.transferfn.TransferFnWithState

Smoothly interpolates a matrix between simulation time steps, with exponential falloff.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number time_constant (allow_None=False, bounds=None, constant=False, default=0.3, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Controls the time scale of the interpolation.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82fc8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c58>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82f70>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82d08>

set_param = <functools.partial object at 0x2b07b1f82c00>

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.DivisiveNormalizeL1(**params)

Bases: imagen.transferfn.TransferFn

TransferFn that divides an array by its L1 norm.

This operation ensures that the sum of the absolute values of the array is equal to the specified norm_value, rescaling each value to make this true. The array is unchanged if the sum of absolute values is zero. For arrays of non-negative values where at least one is non-zero, this operation is equivalent to a divisive sum normalization.

param Number norm_value (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e68>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82ec0>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82f18>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82c00>

set_param = <functools.partial object at 0x2b07b1f82d60>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.DivisiveNormalizeL2(**params)

Bases: imagen.transferfn.TransferFn

TransferFn to divide an array by its Euclidean length (aka its L2 norm).

For a given array interpreted as a flattened vector, keeps the Euclidean length of the vector at a specified norm_value.

param Number norm_value (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82b50>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82f70>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e10>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82c58>

set_param = <functools.partial object at 0x2b07b1f82ba8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.HomeostaticMaxEnt(**params)

Bases: topo.transferfn.TransferFnWithRandomState

Implementation of homeostatic intrinsic plasticity from Jochen Triesch, ICANN 2005, LNCS 3696 pp.65-70.

A sigmoid activation function is adapted automatically to achieve desired average firing rate and approximately exponential distribution of firing rates (for the maximum possible entropy).

Note that this TransferFn has state, so the history of calls to it will affect future behavior. The plastic parameter can be used to disable changes to the state.

Also calculates average activity as useful debugging information, for use with AttributeTrackingTF. Average activity is calculated as an exponential moving average with a smoothing factor (smoothing). For more information see: NIST/SEMATECH e-Handbook of Statistical Methods, Single Exponential Smoothing http://www.itl.nist.gov/div898/handbook/pmc/section4/pmc431.htm

param Parameter b_init (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Additive parameter controlling the exponential.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Parameter random_generator (allow_None=False, constant=False, default=<mtrand.RandomState object at 0x2b07ab473c50>, instantiate=False, pickle_default_value=True, precedence=-1, readonly=False)

Using Numpy’s RandomState instead of random.Random as the former can generate random arrays and more random distributions. See RandomState’s help for more information.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Number mu (allow_None=False, bounds=None, constant=False, default=0.01, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Target average firing rate.

param Number step (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

How often to update the a and b parameters. For instance, step=1 means to update it every time this TF is called; step=2 means to update it every other time.

param Number eta (allow_None=False, bounds=None, constant=False, default=0.0002, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Learning rate for homeostatic plasticity.

param Parameter a_init (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Multiplicative parameter controlling the exponential.

param Boolean time_dependent (allow_None=False, bounds=(0, 1), constant=False, default=False, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether the given time_fn should be used to constrain the results generated.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number smoothing (allow_None=False, bounds=None, constant=False, default=0.9997, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Weighting of previous activity vs. current activity when calculating the average.

param Callable time_fn (allow_None=False, constant=False, default=<Time Time00001>, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Callable used to specify the time that determines the state and return value of the object, if time_dependent=True.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82ec0>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c00>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e68>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82b50>

set_param = <functools.partial object at 0x2b07b1f82cb0>

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.BinaryThreshold(**params)

Bases: imagen.transferfn.TransferFn

Forces all values below a threshold to zero, and above it to 1.0.

param Number threshold (allow_None=False, bounds=None, constant=False, default=0.25, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Decision point for determining binary value.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e10>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82cb0>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82fc8>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82ec0>

set_param = <functools.partial object at 0x2b07b1f82f18>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.ActivityAveragingTF(**params)

Bases: imagen.transferfn.TransferFnWithState

Calculates the average of the input activity.

The average is calculated as an exponential moving average, where the weighting for each older data point decreases exponentially. The degree of weighing for the previous values is expressed as a constant smoothing factor.

The plastic parameter allows the updating of the average values to be disabled temporarily, e.g. while presenting test patterns.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number initial_average (allow_None=False, bounds=None, constant=False, default=0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Starting value for the average activity.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Number step (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

How often to update the average. For instance, step=1 means to update it every time this TF is called; step=2 means to update it every other time.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number smoothing (allow_None=False, bounds=None, constant=False, default=0.9997, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The degree of weighting for the previous average, when calculating the new average.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82c00>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82e10>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82d60>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82e68>

set_param = <functools.partial object at 0x2b07b1f82d08>

state_pop()

Restore the state saved by the most recent state_push call.

state_push()

Save the current state onto a stack, to be restored with state_pop.

Subclasses must implement state_push and state_pop to store state across invocations. The behaviour should be such that after state_pop, the state is restored to what it was at the time when state_push was called.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.ExpLinear(**params)

Bases: imagen.transferfn.TransferFn

Transfer function that is exponential until t from which point it is linear.

param Number a (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The overall scaling of the function

param Number e (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The exponent of the exponetial part of the curve

param Number t2 (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The threshold level at which curve becomes linear

param Number t1 (allow_None=False, bounds=None, constant=False, default=0.5, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The threshold level where function becomes non-zero

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82ba8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82f18>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82d60>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82d08>

set_param = <functools.partial object at 0x2b07b1f82fc8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.DivisiveNormalizeLinf(**params)

Bases: imagen.transferfn.TransferFn

TransferFn to divide an array by its L-infinity norm (i.e. the maximum absolute value of its elements).

For a given array interpreted as a flattened vector, scales the elements divisively so that the maximum absolute value is the specified norm_value.

The L-infinity norm is also known as the divisive infinity norm and Chebyshev norm.

param Number norm_value (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e68>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c58>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b2115100>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b2115470>

set_param = <functools.partial object at 0x2b07b21154c8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.GeneralizedLogistic(**params)

Bases: imagen.transferfn.TransferFn

The generalized logistic curve (Richards’ curve): y = l + (u /(1 + b * exp(-r*(x-2*m))^(1/b))).

The logistic curve is a flexible function for specifying a nonlinear growth curve using five parameters:

• l: the lower asymptote
• u: the upper asymptote minus l
• m: the time of maximum growth
• r: the growth rate
• b: affects near which asymptote maximum growth occurs

From Richards, F.J. (1959), A flexible growth function for empirical use. J. Experimental Botany 10: 290–300. http://en.wikipedia.org/wiki/Generalised_logistic_curve

param Number b (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter which affects near which asymptote maximum growth occurs.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number m (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the time of maximum growth.

param Number l (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the lower asymptote.

param Number r (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the growth rate.

param Number u (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the upper asymptote (upper asymptote minus lower asymptote.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82d08>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82f18>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82fc8>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82d60>

set_param = <functools.partial object at 0x2b07b1f82c00>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.Square(**params)

Bases: imagen.transferfn.TransferFn

Transfer function that applies a squaring nonlinearity.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82d08>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82e68>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82e10>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82f70>

set_param = <functools.partial object at 0x2b07b1f82ec0>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.Sigmoid(**params)

Bases: imagen.transferfn.TransferFn

Sigmoidal (logistic) transfer function: 1/(1+exp-(r*x+k)).

As defined in Jochen Triesch, ICANN 2005, LNCS 3696 pp. 65-70. The parameters control the growth rate (r) and the x position (k) of the exponential.

This function is a special case of the GeneralizedLogistic function, with parameters r=r, l=0, u=1, m=-k/2r, and b=1. See Richards, F.J. (1959), A flexible growth function for empirical use. J. Experimental Botany 10: 290–300, 1959. http://en.wikipedia.org/wiki/Generalised_logistic_curve

param Number k (allow_None=False, bounds=None, constant=False, default=0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the x-postion

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number r (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Parameter controlling the growth rate

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82d08>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82fc8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82c58>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82f18>

set_param = <functools.partial object at 0x2b07b1f82db8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.HalfRectifyAndSquare(**params)

Bases: imagen.transferfn.TransferFn

Transfer function that applies a half-wave rectification (clips at zero) and then squares the values.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number t (allow_None=False, bounds=None, constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The threshold at which output becomes non-zero.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82ec0>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f3f050>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f3f158>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f3f4c8>

set_param = <functools.partial object at 0x2b07b1f3f520>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.NakaRushton(**params)

Bases: imagen.transferfn.TransferFn

Naka-Rushton curve.

From Naka, K. and Rushton, W. (1996), S-potentials from luminosity units in the retina of fish (Cyprinidae). J. Physiology 185:587-599.

The Naka-Rushton curve has been shown to be a good approximation of constrast gain control in cortical neurons. The input of the curve is usually contrast, but under the assumption that the firing rate of a model neuron is directly proportional to the contrast, it can be used as a TransferFn for a Sheet.

The parameter c50 corresponds to the contrast at which the half of the maximal output is reached. For a Sheet TransferFn this translates to the input for which a neuron will respond with activity 0.5.

param Number e (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The exponent of the input x.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number c50 (allow_None=False, bounds=None, constant=False, default=0.1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The input of the neuron at which it responds at half of its maximal firing rate (1.0).

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e68>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82c58>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82c00>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82f70>

set_param = <functools.partial object at 0x2b07b1f82f18>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.Threshold(**params)

Bases: imagen.transferfn.TransferFn

Forces all values below a threshold to zero, and leaves others unchanged.

param Number threshold (allow_None=False, bounds=None, constant=False, default=0.25, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Decision point for determining values to clip.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82e68>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82cb0>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82fc8>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82ec0>

set_param = <functools.partial object at 0x2b07b1f82db8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.DivisiveNormalizeLp(**params)

Bases: imagen.transferfn.TransferFn

TransferFn to divide an array by its Lp-Norm, where p is specified.

For a parameter p and a given array interpreted as a flattened vector, keeps the Lp-norm of the vector at a specified norm_value. Faster versions are provided separately for the typical L1-norm and L2-norm cases. Defaults to be the same as an L2-norm, i.e., DivisiveNormalizeL2.

param Number p (allow_None=False, bounds=None, constant=False, default=2, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param Number norm_value (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b1f82c00>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b1f82e10>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b1f82f70>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b1f82e68>

set_param = <functools.partial object at 0x2b07b1f82f18>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

---

misc Module

Transfer functions with more complex dependencies.

$Id: basic.py 10790 2009-11-21 17:51:33Z antolikjan $

class topo.transferfn.misc.PatternCombine(**params)

Bases: imagen.transferfn.TransferFn

Combine the supplied pattern with one generated using a PatternGenerator.

Useful for operations like adding noise or masking out lesioned items or around the edges of non-rectangular shapes.

param Parameter operator (allow_None=False, constant=False, default=<ufunc ‘multiply’>, instantiate=False, pickle_default_value=True, precedence=0.98, readonly=False)

Binary Numeric function used to combine the two patterns. Any binary Numeric array “ufunc” returning the same type of array as the operands and supporting the reduce operator is allowed here. See topo.pattern.Composite.operator for more details.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param ClassSelector generator (allow_None=False, constant=False, default=<Constant Constant01519>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

Pattern to combine with the supplied matrix.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a0578>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a0680>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a0788>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

operator = <ufunc 'multiply'>

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a0b50>

set_param = <functools.partial object at 0x2b07b28a0ba8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

topo.transferfn.misc.HomeostaticResponse

alias of AdaptiveThreshold

class topo.transferfn.misc.KernelMax(**params)

Bases: imagen.transferfn.TransferFn

Replaces the given matrix with a kernel function centered around the maximum value.

This operation is usually part of the Kohonen SOM algorithm, and approximates a series of lateral interactions resulting in a single activity bubble.

The radius of the kernel (i.e. the surround) is specified by the parameter ‘radius’, which should be set before using __call__. The shape of the surround is determined by the neighborhood_kernel_generator, and can be any PatternGenerator instance, or any function accepting bounds, density, radius, and height to return a kernel matrix.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number density (allow_None=False, bounds=(0, None), constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Density of the Sheet whose matrix we act on, for use in converting from matrix to Sheet coordinates.

param Number crop_radius_multiplier (allow_None=False, bounds=None, constant=False, default=3.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Factor by which the radius should be multiplied, when deciding how far from the winner to keep evaluating the kernel.

param Number kernel_radius (allow_None=False, bounds=(0, None), constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Kernel radius in Sheet coordinates.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param ClassSelector neighborhood_kernel_generator (allow_None=False, constant=False, default=<Gaussian Gaussian01520>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

Neighborhood function

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a06d8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a08e8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a0890>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a02b8>

set_param = <functools.partial object at 0x2b07b28a0310>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.misc.AdaptiveThreshold(**params)

Bases: imagen.transferfn.TransferFnWithState

Adapts the parameters of a linear threshold function to maintain a constant desired average activity. Defined in:

Jean-Luc R. Stevens, Judith S. Law, Jan Antolik, and James A. Bednar. Mechanisms for stable, robust, and adaptive development of orientation maps in the primary visual cortex. Journal of Neuroscience 33:15747-15766, 2013. http://dx.doi.org/10.1523/JNEUROSCI.1037-13.2013

param Number smoothing (allow_None=False, bounds=None, constant=False, default=0.991, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Weighting of previous activity vs. current activity when calculating the average activity.

param Number t_init (allow_None=False, bounds=None, constant=False, default=0.15, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Initial value of the threshold value t.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number learning_rate (allow_None=False, bounds=None, constant=False, default=0.01, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Learning rate for homeostatic plasticity.

param Number period (allow_None=False, bounds=None, constant=True, default=1.0, inclusive_bounds=(True, True), instantiate=True, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

How often the threshold should be adjusted. If the period is 0, the threshold is adjusted continuously, each time this TransferFn is called. For nonzero periods, adjustments occur only the first time this TransferFn is called after topo.sim.time() reaches an integer multiple of the period. For example, if period is 2.5 and the TransferFn is evaluated every 0.05 simulation time units, the threshold will be adjusted at times 2.55, 5.05, 7.55, etc.

param Number noise_magnitude (allow_None=False, bounds=None, constant=False, default=0.1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The magnitude of the additive noise to apply to the t_init parameter at initialization.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Integer seed (allow_None=False, bounds=None, constant=False, default=42, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Random seed used to control the initial randomized t values.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Boolean randomized_init (allow_None=False, bounds=(0, 1), constant=False, default=False, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether to randomize the initial t parameter.

param Number target_activity (allow_None=False, bounds=None, constant=False, default=0.024, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The target average activity.

param Number linear_slope (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

Slope of the linear portion above threshold.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a0998>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a0a48>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a0aa0>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a0680>

set_param = <functools.partial object at 0x2b07b28a0628>

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.misc.HalfRectify(**params)

Bases: imagen.transferfn.TransferFn

Transfer function that applies a half-wave rectification (clips at zero)

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Number noise_magnitude (allow_None=False, bounds=None, constant=False, default=0.1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The magnitude of the additive noise to apply to the t_init parameter at initialization.

param Boolean randomized_init (allow_None=False, bounds=(0, 1), constant=False, default=False, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether to randomize the initial t parameter.

param Number gain (allow_None=False, bounds=None, constant=False, default=1.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The neuronal gain

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Number t_init (allow_None=False, bounds=None, constant=False, default=0.0, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The initial value of threshold at which output becomes non-zero..

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a0940>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a0af8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a0310>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a0db8>

set_param = <functools.partial object at 0x2b07b28a04c8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.misc.TemporalScatter(**params)

Bases: imagen.transferfn.TransferFnWithState

Scatter values across time using a specified distribution, discretized into a symmetric interval around zero. This class is still work in progress as part of the TCAL model.

As no notion of time exists at the level of transfer functions (state is changes according to call count), this class assumes a fixed, ‘clocked’ timestep exists between subsequent calls.

Internally, an activity buffer is computed with a depth corresponding to the number of timestep intervals with the stated span value.

Note that the transfer function only has the power to delay output. Therefore the central peak of a unimodal, zero-mean distribution will occur after the time ‘span’ has elapsed.

In addition it is very important to view the depth map using the view_depth_map method: if the majority of the values generated by the distribution are outside the chosen span, values smaller and larger than the span will be lumped into the first and last bins respectively, distorting the shape of the intended distribution.

param Number span (allow_None=True, bounds=(0, None), constant=False, default=120, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The input distribution is expected to be continuous and may be unbounded. For instance, a Gaussian distribution may generate sample values that are unbounded in both the positive and negative direction. The span parameter determines the size of the (zero-centered) interval which is binned.

param Number timestep (allow_None=False, bounds=None, constant=False, default=5, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

The timestep interval in milliseconds. This value is used to compute the depth and sample the supplied distribution. Note that value must be specified some some extenal source and there is no way to ensure that subsequent calls are regular with the stated interval.

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param ClassSelector distribution (allow_None=False, constant=False, default=<GaussianRandom GaussianRandom01521>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

The pattern generator that defines the scatter distribution in milliseconds. Any random distribution may be used e.g. UniformRandom or GaussianRandom. Note that the discretized binning with the given ‘span’ is zero-centered. In other words, even if a distribution is not-symmetric (i.e. skewed), binning will occur around a symmetric interval around zero with a total extent given by the span.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

depth

The depth of the activity buffer.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a0b50>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a0418>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a0680>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a0a48>

set_param = <functools.partial object at 0x2b07b28a0e10>

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

view_depth_map(mode='both')

Visualize the depth map using holoviews, including distribution histograms.

Mode may be one of ‘discrete’, ‘raw’ or ‘both’:

• The ‘discrete’ mode presents the depth map used by TemporalScatter, showing the latency at which TemporalScatter will propagate the input i.e. discrete, positive latencies in milliseconds.
• The ‘raw’ mode shows the continuous distribution before discretization. This is typically a zero-mean, zero-centered distribution i.e a continuous, zero-centered latency distribution.
• Both presents both of the above types together (default).

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.misc.AttributeTrackingTF(**params)

Bases: imagen.transferfn.TransferFnWithState

Keeps track of attributes of a specified Parameterized over time, for analysis or plotting.

Useful objects to track include sheets (e.g. “topo.sim[‘V1’]”), projections (“topo.sim[‘V1’].projections[‘LateralInhibitory’]”), or an output_function.

Any attribute whose value is a matrix the same size as the activity matrix can be tracked. Only specified units within this matrix will be tracked.

If no object is specified, this function will keep track of the incoming activity over time.

The results are stored in a dictionary named ‘values’, as (time, value) pairs indexed by the parameter name and unit. For instance, if the value of attribute ‘x’ is v for unit (0.0,0.0) at time t, values[‘x’][(0.0,0.0)]=(t,v).

Updating of the tracked values can be disabled temporarily using the plastic parameter.

param List init_keys (allow_None=False, bounds=(0, None), constant=True, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of item key labels for metadata that that must be supplied to the initialize method before the TransferFn may be used.

param List attrib_names (allow_None=False, bounds=(0, None), constant=False, default=[], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

List of names of the function object’s parameters that should be stored.

param Parameter object (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Parameterized instance whose parameters will be tracked. If this parameter’s value is a string, it will be evaluated first (by calling Python’s eval() function). This feature is designed to allow circular references, so that the TF can track the object that owns it, without causing problems for recursive traversal (as for script_repr()).

param Boolean plastic (allow_None=False, bounds=(0, 1), constant=False, default=True, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

Whether or not to update the internal state on each call. Allows plasticity to be temporarily turned off (e.g for analysis purposes).

param Number step (allow_None=False, bounds=None, constant=False, default=1, inclusive_bounds=(True, True), instantiate=False, pickle_default_value=True, precedence=None, readonly=False, time_dependent=True, time_fn=<Time Time00001>)

How often to update the tracked values. For instance, step=1 means to update them every time this TF is called; step=2 means to update them every other time.

param List units (allow_None=False, bounds=(0, None), constant=False, default=[(0.0, 0.0)], instantiate=True, pickle_default_value=True, precedence=None, readonly=False)

Sheet coordinates of the unit(s) for which parameter values will be stored.

param Parameter norm_value (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

param Parameter coordframe (allow_None=True, constant=False, default=None, instantiate=False, pickle_default_value=True, precedence=None, readonly=False)

SheetCoordinateSystem to use to convert the position into matrix coordinates. If this parameter’s value is a string, it will be evaluated first(by calling Python’s eval() function). This feature is designed to allow circular references, so that the TF can track the object that owns it, without causing problems for recursive traversal (as for script_repr()).

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b28a0470>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b28a0db8>

initialize(**kwargs)

Transfer functions may need additional information before the supplied numpy array can be modified in place. For instance, transfer functions may have state which needs to be allocated in memory with a certain size. In other cases, the transfer function may need to know about the coordinate system associated with the input data.

inspect_value = <functools.partial object at 0x2b07b28a06d8>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

override_plasticity_state(new_plasticity_state)

Temporarily disable plasticity of internal state.

This function should be implemented by all subclasses so that after a call, the output should always be the same for any given input pattern, and no call should have any effect that persists after restore_plasticity_state() is called.

By default, simply saves a copy of the ‘plastic’ parameter to an internal stack (so that it can be restored by restore_plasticity_state()), and then sets the plastic parameter to the given value (True or False).

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

restore_plasticity_state()

Re-enable plasticity of internal state after an override_plasticity_state call.

This function should be implemented by all subclasses to remove the effect of the most recent override_plasticity_state call, i.e. to reenable changes to the internal state, without any lasting effect from the time during which plasticity was disabled.

By default, simply restores the last saved value of the ‘plastic’ parameter.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b28a0ba8>

set_param = <functools.partial object at 0x2b07b28a0940>

state_pop()

Restore the state saved by the most recent state_push call.

state_push()

Save the current state onto a stack, to be restored with state_pop.

Subclasses must implement state_push and state_pop to store state across invocations. The behaviour should be such that after state_pop, the state is restored to what it was at the time when state_push was called.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

---

optimized Module

Output functions (see basic.py) and projection-level output functions (see projfn.py) written in C to optimize performance.

Requires the weave package; without it unoptimized versions are used.

class topo.transferfn.optimized.CFPOF_DivisiveNormalizeL1_opt(**params)

Bases: topo.base.cf.CFPOutputFn

Performs divisive normalization of the weights of all cfs. Intended to be equivalent to, but faster than, CFPOF_DivisiveNormalizeL1.

param ClassSelector single_cf_fn (allow_None=False, constant=True, default=<DivisiveNormalizeL1 DivisiveNormalizeL101517>, instantiate=False, is_instance=True, pickle_default_value=True, precedence=None, readonly=True)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b2a11c00>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b2a11e10>

inspect_value = <functools.partial object at 0x2b07b2a11f18>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b2a11e68>

set_param = <functools.partial object at 0x2b07b2a11fc8>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.optimized.CFPOF_DivisiveNormalizeL1(**params)

Bases: topo.base.cf.CFPOutputFn

Non-optimized version of CFPOF_DivisiveNormalizeL1_opt.

Same as CFPOF_Plugin(single_cf_fn=DivisiveNormalizeL1()), except that it supports joint normalization using the norm_total property of ConnectionField.

param ClassSelector single_cf_fn (allow_None=False, constant=True, default=<DivisiveNormalizeL1 DivisiveNormalizeL101518>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b2a11d08>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b2a11c00>

inspect_value = <functools.partial object at 0x2b07b2a11d60>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b2a11c58>

set_param = <functools.partial object at 0x2b07b2a11f70>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

topo.transferfn.optimized.DivisiveNormalizeL1_opt

alias of DivisiveNormalizeL1

---

projfn Module

Output functions (see basic.py) that apply to a whole Projection. For example, for a CFProjection this involves iterating through all the CFs and applying an output function to each set of weights in turn.

class topo.transferfn.projfn.CFPOF_Plugin(**params)

Bases: topo.base.cf.CFPOutputFn

Applies the specified single_cf_fn to each CF in the CFProjection for which the mask is nonzero.

param ClassSelector single_cf_fn (allow_None=False, constant=False, default=<IdentityTF IdentityTF01277>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

Accepts a TransferFn that will be applied to each CF individually.

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b239d4c8>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b239d260>

inspect_value = <functools.partial object at 0x2b07b239d578>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b239d520>

set_param = <functools.partial object at 0x2b07b239d680>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

class topo.transferfn.projfn.CFPOF_Identity(**params)

Bases: topo.base.cf.CFPOutputFn

CFPOutputFn that leaves the CFs unchanged.

Must never be changed or subclassed, because it might never be called. (I.e., it could simply be tested for and skipped.)

param ClassSelector single_cf_fn (allow_None=False, constant=True, default=<IdentityTF IdentityTF01278>, instantiate=True, is_instance=True, pickle_default_value=True, precedence=None, readonly=False)

debug(msg, *args, **kw)

Print msg merged with args as a debugging statement.

See Python’s logging module for details of message formatting.

defaults()

Return {parameter_name:parameter.default} for all non-constant Parameters.

Note that a Parameter for which instantiate==True has its default instantiated.

force_new_dynamic_value = <functools.partial object at 0x2b07b239d368>

get_param_values(onlychanged=False)

Return a list of name,value pairs for all Parameters of this object.

If onlychanged is True, will only return values that are not equal to the default value.

get_value_generator = <functools.partial object at 0x2b07b239d2b8>

inspect_value = <functools.partial object at 0x2b07b239d3c0>

message(msg, *args, **kw)

Print msg merged with args as a message.

See Python’s logging module for details of message formatting.

classmethod params(parameter_name=None)

Return the Parameters of this class as the dictionary {name: parameter_object}

Includes Parameters from this class and its superclasses.

pprint(imports=None, prefix=' ', unknown_value='<?>', qualify=False, separator='')

(Experimental) Pretty printed representation that may be evaluated with eval. See pprint() function for more details.

classmethod print_param_defaults()

Print the default values of all cls’s Parameters.

print_param_values()

Print the values of all this object’s Parameters.

script_repr(imports=, []prefix=' ')

Variant of __repr__ designed for generating a runnable script.

classmethod set_default(param_name, value)

Set the default value of param_name.

Equivalent to setting param_name on the class.

set_dynamic_time_fn = <functools.partial object at 0x2b07b239d260>

set_param = <functools.partial object at 0x2b07b239d470>

state_pop()

Restore the most recently saved state.

See state_push() for more details.

state_push()

Save this instance’s state.

For Parameterized instances, this includes the state of dynamically generated values.

Subclasses that maintain short-term state should additionally save and restore that state using state_push() and state_pop().

Generally, this method is used by operations that need to test something without permanently altering the objects’ state.

verbose(msg, *args, **kw)

Print msg merged with args as a verbose message.

See Python’s logging module for details of message formatting.

warning(msg, *args, **kw)

Print msg merged with args as a warning, unless module variable warnings_as_exceptions is True, then raise an Exception containing the arguments.

See Python’s logging module for details of message formatting.

---