bluepyemodel.evaluation.efel_feature_bpem

Class eFELFeatureBPEM

Classes

DendFitFeature(name[, efel_feature_name, ...])

Fit to back propagation feature

DendFitMultiProtocolsFeature(name[, ...])

Fit across apical dendrite using multiple protocols.

eFELFeatureBPEM(name[, efel_feature_name, ...])

eFEL feature extra
class DendFitFeature(name, efel_feature_name=None, recording_names=None, stim_start=None, stim_end=None, exp_mean=None, exp_std=None, threshold=None, stimulus_current=None, comment='', interp_step=None, double_settings=None, int_settings=None, string_settings=None, decay=None, linear=None, weight=1.0)

Bases: eFELFeatureBPEM

Fit to back propagation feature

To use this class:

  • have “dendrite_backpropagation_fit” as the efeature name

  • have “maximum_voltage_from_voltagebase” as the efel_feature_name

  • have keys in recording names matching the distance from soma, and “” for soma, e.g.

    {“”: “soma.v”, “50”: “dend50.v”, “100”: “dend100.v”, “150”: “dend150.v”}

  • have appropriate recordings in protocols

Constructor

calculate_feature(responses, raise_warnings=False)

Calculate feature value

calculate_score(responses, trace_check=False)

Calculate score. bpo and non-bpo feature should use calulate_feature

fit(distances, values)

Fit back propagation

get_distances_feature_values(responses, raise_warnings=False)

Compute feature at each distance, and return distances and feature values.

class DendFitMultiProtocolsFeature(name, efel_feature_name=None, recording_names=None, stim_start=None, stim_end=None, exp_mean=None, exp_std=None, threshold=None, stimulus_current=None, comment='', interp_step=None, double_settings=None, int_settings=None, string_settings=None, decay=None, linear=None, weight=1.0)

Bases: DendFitFeature

Fit across apical dendrite using multiple protocols.

Attention! Since this feature depends on multiple protocols, the stimulus_current passed can be wrong for some of them, and in such a case, efel features depending on stimulus_current should not be used.

To use this class:
  • have a protocol_name with distances in brackets in it

    e.g. LocalInjectionIDrestapic[050,100,150,200]_100

  • same with recording_name

    e.g. apical[055,080,110,200,340].v

  • have corresponding protocols with normal names under “protocols” in config

    e.g. LocalInjectionIDrestapic050_100, LocalInjectionIDrestapic100_100, etc.

  • have a soma protocol under “protocols” with same ecode in config

    e.g. IDrest_100

Constructor

class eFELFeatureBPEM(name, efel_feature_name=None, recording_names=None, stim_start=None, stim_end=None, exp_mean=None, exp_std=None, threshold=None, stimulus_current=None, comment='', interp_step=None, double_settings=None, int_settings=None, string_settings=None, weight=1.0)

Bases: eFELFeature

eFEL feature extra

Constructor

Parameters:

  • name (str) – name of the eFELFeature object

  • efel_feature_name (str) – name of the eFeature in the eFEL library (ex: ‘AP1_peak’)

  • recording_names (dict) – eFEL features can accept several recordings as input

  • stim_start (float) – stimulation start time (ms)

  • stim_end (float) – stimulation end time (ms)

  • exp_mean (float) – experimental mean of this eFeature

  • exp_std (float) – experimental standard deviation of this eFeature

  • threshold (float) – spike detection threshold (mV)

  • comment (str) – comment

  • weight (float) – weight of the efeature. Basically multiplies the score of the efeature by this value.

calculate_bpo_feature(responses)

Return internal feature which is directly passed as a response

calculate_bpo_score(responses)

Return score for bpo feature

calculate_feature(responses, raise_warnings=False)

Calculate feature value

calculate_score(responses, trace_check=False)

Calculate the score

calulate_score_(responses)

Calculate the score for non-bpo feature