Preprocessing

Any transformation of the data matrix that is not a tool. Other than tools, preprocessing steps usually don’t return an easily interpretable annotation, but perform a basic transformation on the data matrix.

Basic preprocessing

preprocessing.encode	Encode categoricals of an AnnData object.
preprocessing.pca	Computes a principal component analysis.
preprocessing.regress_out	Regress out (mostly) unwanted sources of variation.
preprocessing.subsample	Subsample to a fraction of the number of observations.
preprocessing.balanced_sample	Balancing groups in the dataset.
preprocessing.highly_variable_features	Annotate highly variable features.
preprocessing.winsorize	Returns a Winsorized version of the input array.
preprocessing.clip_quantile	Clips (limits) features.
preprocessing.summarize_measurements	Summarizes numerical measurements into minimum, maximum and average values.

Quality control

preprocessing.qc_metrics	Calculates various quality control metrics.
preprocessing.qc_lab_measurements	Examines lab measurements for reference ranges and outliers.
preprocessing.mcar_test	Statistical hypothesis test for Missing Completely At Random (MCAR).
preprocessing.detect_bias	Detects biases in the data using feature correlations, standardized mean differences, and feature importances.

Imputation

preprocessing.explicit_impute	Replaces all missing values in all columns or a subset of columns specified by the user with the passed replacement value.
preprocessing.simple_impute	Impute missing values in numerical data using mean/median/most frequent imputation.
preprocessing.knn_impute	Imputes missing values in the input AnnData object using K-nearest neighbor imputation.
preprocessing.miss_forest_impute	Impute data using the MissForest strategy.
preprocessing.mice_forest_impute	Impute data using the miceforest.

Normalization

preprocessing.log_norm	Apply log normalization.
preprocessing.maxabs_norm	Apply max-abs normalization.
preprocessing.minmax_norm	Apply min-max normalization.
preprocessing.power_norm	Apply power transformation normalization.
preprocessing.quantile_norm	Apply quantile normalization.
preprocessing.robust_scale_norm	Apply robust scaling normalization.
preprocessing.scale_norm	Apply scaling normalization.
preprocessing.offset_negative_values	Offsets negative values into positive ones with the lowest negative value becoming 0.

Dataset Shift Correction

Partially overlaps with dataset integration. Note that a simple batch correction method is available via pp.regress_out().

preprocessing.combat

ComBat function for batch effect correction [JLR06], [LJP+17], [Ped12].

Neighbors

preprocessing.neighbors

Compute a neighborhood graph of observations [MHM18].