mcmodels.models.VoxelModel

class mcmodels.models.VoxelModel(source_voxels, kernel='linear', degree=3, coef0=1, gamma=None, kernel_params=None)

Voxel-scale interpolation model for mesoscale connectivity.

Parameters:

source_voxels : array-like, shape=(n_voxels, 3)

List of voxel coordinates at which to interpolate.

See also

NadarayaWatson

References

Knox et al. ‘High resolution data-driven model of the mouse connectome’.

bioRxiv 293019; doi: https://doi.org/10.1101/293019

Examples

>>> from mcmodels.core import VoxelModelCache
>>> from mcmodels.models import VoxelModel
>>> cache = VoxelModelCache
>>> # get cortical experiment data
>>> cortex_data = cache.get_experiment_data(injection_structure_ids=[315])
>>> source_voxels = cortex_data.source_mask.coordinates
>>> reg = VoxelModel(source_voxels)
>>> reg.fit((cortex_data.centroids, cortex_data.injections))
VoxelModel(source_voxels=array([[ ... ]]))

Attributes:

nodes

Return model nodes (data).

weights

Return model weights.

Methods

fit(self, X, y[, sample_weight])	Fit Voxel Model.
get_params(self[, deep])	Get parameters for this estimator.
get_weights(self)	Overwrite of NadarayaWatson.get_weights.
predict(self, X)	Predict projection volumes given injection volumes.
score(self, X, y[, sample_weight])	Returns the coefficient of determination R^2 of the prediction.
set_params(self, \*\*params)	Set the parameters of this estimator.

__init__(self, source_voxels, kernel='linear', degree=3, coef0=1, gamma=None, kernel_params=None)

Methods

__init__(self, source_voxels[, kernel, …])
fit(self, X, y[, sample_weight])	Fit Voxel Model.
get_params(self[, deep])	Get parameters for this estimator.
get_weights(self)	Overwrite of NadarayaWatson.get_weights.
predict(self, X)	Predict projection volumes given injection volumes.
score(self, X, y[, sample_weight])	Returns the coefficient of determination R^2 of the prediction.
set_params(self, \*\*params)	Set the parameters of this estimator.

fit(self, X, y, sample_weight=None)

Fit Voxel Model.

NOTE : X is a concatenation (column wise) of the injection centroid

coordinates and the injection volumes. This choice was made to be consistent with the sklearn.core.BaseEstimator fit and predict schema

Parameters:

X : {array-like, sparse matrix}, shape=(n_exps, 3+n_source_voxels)

Centroid coordinates concatenated with the injection density for each training experiment.

y : {array-like, sparse matrix}, shape=(n_exps, n_target_voxels)

Normalized projection density for each training experiment

Returns:

self : returns an instance of self.

get_weights(self)

Overwrite of NadarayaWatson.get_weights.

nodes

Return model nodes (data).

predict(self, X)

Predict projection volumes given injection volumes.

NOTE : X is a concatenation (column wise) of the injection centroid

coordinates and the injection volumes. This choice was made to be consistent with the sklearn.core.BaseEstimator fit and predict schema

Parameters:

X : {array-like, sparse matrix}, shape=(n_exps, 3+n_source_voxels)

Centroid coordinates concatenated with the injection density for each test experiment.

Returns:

C : array, shape=(X.shape[0], y.shape[1])

Predicted normalized projection densities.

weights

Return model weights.