pegasus.select_features

pegasus.select_features(data, features='highly_variable_features', standardize=True, max_value=10.0)[source]

Subset the features and store the resulting matrix in dense format in data.uns with ‘_tmp_fmat_’ prefix, with the option of standardization and truncating based on max_value. ‘_tmp_fmat_*’ will be removed before writing out the disk. :type data: Union[MultimodalData, UnimodalData] :param data: Annotated data matrix with rows for cells and columns for genes. :type data: pegasusio.MultimodalData :type features: str :param features: a keyword in data.var, which refers to a boolean array. If None, all features will be selected. :type features: str, optional, default: highly_variable_features. :type standardize: bool :param standardize: Whether to scale the data to unit variance and zero mean. :type standardize: bool, optional, default: True. :type max_value: float :param max_value: The threshold to truncate data after scaling. If None, do not truncate. :type max_value: float, optional, default: 10.

Return type

str

Returns

  • keyword (str) – The keyword in data.uns referring to the features selected.

  • Update data.uns if needed –

    • data.uns[keyword]: A submatrix of the data containing features selected.

Examples

>>> pg.select_features(data)