atlas_protocol_scripts.pp.is_outlier#

atlas_protocol_scripts.pp.is_outlier(adata, metric_col, *, groupby=None, n_mads=5)#

Detect outliers by median absolute deviation (MAD).

Adapted from https://www.sc-best-practices.org/preprocessing_visualization/quality_control.html#motivation

Parameters

  • adata (AnnData) – AnnData object

  • metric_col (str) – column in adata.obs to consider

  • groupby (Optional[str] (default: None)) – grouping variable. If specified, outliers will be determined for each value of this grouping variable separately. E.g. dataset.

  • n_mads (float (default: 5)) – label points that are outside of median +/- nmads * MAD.

Return type

Series