episcanpy.pp.select_var_feature(adata, max_score=0.5, nb_features=None, show=True, copy=False)

This function computes a variability score to rank the most variable features across all cells. Then it selects the most variable features according to either a specified number of features (nb_features) or a minimum variance score (min_score).

adata : adata object

max_score : max threshold of the variability score to retain features,

0 is the score of the most variable features and 0.5 is the score of the least variable features. : where

nb_features : default value is None, if specify it will select a the top most variable features.

the nb_features is larger than the total number of feature, it filters based on the min_score argument : if

show : default value True, it will plot the distribution of var.

copy : return a new adata object if copy == True.


Depending on copy, returns a new AnnData object or overwrite the input