Sparse-Optimization-Toolbox 0.1.0

Package for Sparse optimization

An sparse opyimization toolbox contains test data generation and network reasoning

Test Data Generation

Import

from sparsetools import matCreater

Data generation

matCreater.matCreater(tfLen=10, sampleNums=200, geneNums=2000, normalLoc=0, normalScale=0.1)

Parameter	Type	Explanation
tfLen	int	The numbers of transcribe factors
sampleNums	int	The numbers of transcribe samples
geneNums	int	The numbers of target genes
normalLoc	float: recommond use 0	Mean value of Gaussian noise
normalScale	float	Variance of Gaussian noise

Return:

Parameter	Type	Shapes	Explanation
W_d	np.array	(tfLen, sampleNums)	Over complete dictionary
zNetwork	np.array	(geneNums, tfLen)	Sparse matrix
xTargetGene	np.array	(geneNums,sampleNums)	Target

Network reasoning

from sparsetools import Optimization
Optimization.voting(expre, tf_names, gene_names)

Parameter	Type	Explanation
expre	np.array	The expresion matrix of genes
tf_names	np.array	Names of tf
gene_names	np.array	Name of all genes(including TF)

Return:

The moderation network result matrix of the voting algorithm, while the results of the independent algorithm are named with the algorithm name and stored locally.