splicejac.tools.regr_method_sens
test the sensitivity to the regression method and parameters
Module Contents
Functions
|
Construct the coefficient distribution for the gene-gene interaction matrix G |
|
Compare methods for gene-gene interaction parameter regression |
- splicejac.tools.regr_method_sens.coeff_dist(G, x=np.logspace(- 5, 0, num=100))
Construct the coefficient distribution for the gene-gene interaction matrix G
- G: ~numpy.ndarray
interaction matrix
- x: ~numpy.ndarray (default: numpy.logspace(-5, 0, num=100))
vector of thresholds to group matrix coefficients
- y: ~numpy.ndarray
The distribution of coefficients
- splicejac.tools.regr_method_sens.regr_method_sens(adata, alpha_ridge=np.array([0.01, 0.1, 1, 10, 100]), alpha_lasso=np.array([0.0001, 0.001, 0.01, 0.1, 1]))
Compare methods for gene-gene interaction parameter regression Results are stored in adata.uns[‘method_sens’] and adata.uns[‘sens_coeff’]
- adata: ~anndata.AnnData
count matrix
- alpha_ridge: ~numpy.ndarray (default: numpy.array([0.01, 0.1, 1, 10, 100]))
array of shrinkage coefficients to test for Ridge regression
- alpha_lasso: ~numpy.ndarray (default: numpy.array([0.0001, 0.001, 0.01, 0.1, 1]))
array of shrinkage coefficients to test for Lasso regression
None