dnnbrain.brain.core module

class dnnbrain.brain.core.BrainDecoder(brain_activ=None, map_type=None, estimator=None, cv=5, scoring=None)

Bases: object

Decode brain activation to DNN activation or behavior data.

decode_behavior(beh_data)

Decode brain activation to behavior data.

Parameters

beh_data (ndarray) – Behavior data with shape as (n_stim, n_beh).

Returns

decode_dict – It depends on map type.

map type

key

value

uv

score

If estimator type is correlation, it’s an array with shape as
(n_beh,).
Each element is the maximal pearson r among all measurements
correlating to the corresponding behavior.
If estimator type is regressor or classifier, it’s an array
with shape as (n_beh, cv).
Each row contains scores of each cross validation fold, when
using the measurement at the maximal location to predict the
corresponding behavior.

location

An array with shape as (n_beh,).
Each element is a location of the measurement which makes the
maximal score.

model

An array with shape as (n_beh,).
Each element is a model fitted by the measurement at the
maximal location and the corresponding behavior.
Note: not exist when estimator type is correlation

conf_m

An array with shape as (n_beh, cv).
Each row contains confusion matrices (n_label, n_label) of
each cross validation fold, when using the measurement at the
maximal location to predict the corresponding behavior.
Note: only exists when estimator type is classifier

mv

score

An array with shape as (n_beh, cv).
Each row contains scores of each cross validation fold, when
using all measurements to predict the corresponding behavior.

model

An array with shape as (n_beh,).
Each element is a model fitted by all measurements and the
corresponding behavior.

conf_m

An array with shape as (n_beh, cv).
Each row contains confusion matrices (n_label, n_label) of
each cross validation fold, when using all measurements to
predict the corresponding behavior.
Note: only exists when estimator type is classifier

Return type

dict

decode_dnn(dnn_activ)

Decode brain activation to DNN activation.

Parameters

dnn_activ (Activation) – DNN activation.

Returns

decode_dict – It depends on map type.

Map type

First key

First value

Second key

Second value

uv

layer

(str)

score

If estimator type is correlation, it’s an array with shape
as (n_chn, n_row, n_col) of max scores.
If estimator type is regressor, it’s an array with shape
as (n_chn, n_row, n_col, cv).
The forth dimension contains scores of each cross
validation fold of the max scores.

location

An array with shape as (n_chn, n_row, n_col).
Locations of measurement indicators with max scores.

model

An array with shape as (n_chn, n_row, n_col).
fitted models of the max scores.
Note: only exists when model is regressor

mv

layer

(str)

score

An array with shape as (n_chn, n_row, n_col, cv).
The forth dimension contains scores of each
cross validation fold at each unit.

model

An array with shape as (n_chn, n_row, n_col).
Each element is a model fitted at the corresponding unit.

Return type

dict

set_activ(brain_activ)

Set brain activation

Parameters

brain_activ (ndarray) – Brain activation with shape as (n_vol, n_meas). For voxel-wise, n_meas is the number of voxels. For ROI-wise, n_meas is the number of ROIs.

set_mapper(map_type, estimator, cv, scoring)

Set mapper parameters.

Parameters

  • map_type (str) – choices=(uv, mv). uv: univariate mapping. mv: multivariate mapping.

  • estimator (str | sklearn estimator or pipeline) –

    If is str, it is a name of a estimator used to do mapping. If is ‘corr’, it just uses correlation rather than prediction.

    And the map_type must be ‘uv’.

  • cv (int) – the number of cross validation folds.

  • scoring (str or callable) – the method to evaluate the predictions on the test set.

class dnnbrain.brain.core.BrainEncoder(brain_activ=None, map_type=None, estimator=None, cv=5, scoring=None)

Bases: object

Encode DNN activation or behavior data to brain activation.

encode_behavior(beh_data)

Encode behavior data to brain activation.

Parameters

beh_data (ndarray) – Behavior data with shape as (n_stim, n_beh).

Returns

encode_dict – It depends on map type.

Map type

First key

First value

Second key

Second value

uv

layer

(str)

score

If estimator type is correlation, it’s an array with shape
as (n_meas,) of max scores.
If estimator type is regressor, it’s an array with shape as
(n_meas, cv).
The second dimension contains scores of each
cross validation fold at maximal location.

location

An array with shape as (n_meas,).
Max locations of the max scores.

model

An array with shape as (n_meas,).
Fitted models of the max scores.
Note: only exists when model is regressor.

mv

layer

(str)

score

An array with shape as (n_meas, cv).
The second dimension contains scores of each cross
validation fold.

model

An array with shape as (n_meas,).
Each element is a model fitted at the corresponding
measurement.

Return type

dict

encode_dnn(dnn_activ, iter_axis=None)

Encode DNN activation to brain activation.

Parameters

  • dnn_activ (Activation) – DNN activation.

  • iter_axis (None or str) –

    Iterate along the specified axis. Different map types have different operation.

    map type

    iter_axis

    description

    uv

    channel

    Summarize the maximal prediction score for each channel

    row_col

    Summarize the maximal prediction score for each position (row_idx, col_idx)

    None

    Summarize the maximal prediction score for the whole layer

    mv

    channel

    Multivariate prediction using all units in each channel

    row_col

    Multivariate prediction using all units in each position (row_idx, col_idx)

    None

    Multivariate prediction using all units in the whole layer

Returns

encode_dict – It depends on map type.

Map type

First key

First value

Second key

Second value

uv

layer

(str)

score

If estimator type is correlation, it’s an array with shape
as (n_iter, n_meas).
Each element is the maximal pearson r among all features at
corresponding iteration correlating to the corresponding
measurement.
If estimator type is regressor, it’s an array with shape as
(n_iter, n_meas, cv).
For each iteration and measurement, the third axis contains
scores of each cross validation folds, when using the
feature with maximal score to predict the corresponding
measurement.

location

An array with shape as (n_iter, n_meas, 3)
Max locations of the max scores, the size 3 of the third
dimension means channel, row and column respectively.

model

An array with shape as (n_iter, n_meas).
Fitted models of the max scores.
Note: only exists when estimator type is regressor

mv

layer

(str)

score

A array with shape as (n_iter, n_meas, cv).
The third dimension means scores of each cross validation
folds at each iteration and measurement.

model

A array with shape as (n_iter, n_meas).
Each element is a model fitted at the corresponding
iteration and measurement.

Return type

dict

set_activ(brain_activ)

Set brain activation

Parameters

brain_activ (ndarray) – Brain activation with shape as (n_vol, n_meas). For voxel-wise, n_meas is the number of voxels. For ROI-wise, n_meas is the number of ROIs.

set_mapper(map_type, estimator, cv, scoring)

Set UnivariateMapping or MultivariateMapping

Parameters

  • map_type (str) – choices=(uv, mv) uv: univariate mapping mv: multivariate mapping

  • estimator (str | sklearn estimator or pipeline) –

    If is str, it is a name of a estimator used to do mapping. If is ‘corr’, it just uses correlation rather than prediction.

    And the map_type must be ‘uv’.

    NOTE: The estimator type can only be regressor or correlation

  • cv (int) – the number of cross validation folds.

  • scoring (str or callable) – the method to evaluate the predictions on the test set.

class dnnbrain.brain.core.ROI(rois=None, value=None)

Bases: object

A class used to encapsulate and manipulate ROI data of brain

delete(rois)

Delete data according to ROI names

Parameters

rois (str, list) – ROI names.

get(rois)

Get data according to ROI names

Parameters

rois (str, list) – ROI names.

Returns

arr – ROI data with shape as (n_vol, n_roi).

Return type

ndarray

load(fname, rois=None)

Load from ROI file.

Parameters

  • fname (str) – File name with suffix as .roi.h5

  • rois (str, list) – ROI names of interest.

save(fname)

Save to ROI file.

Parameters

fname (str) – File name with suffix as .roi.h5

set(rois, value, index=None)

Set ROI data with names

Parameters

  • rois (str, list) – ROI names.

  • value (ndarray) – ROI data.

  • index (int) – The position where the data is set.