yasc.preprocessing.rf_fill_missing¶

yasc.preprocessing.rf_fill_missing(data, column, target=None, n_digits=0, **kwargs)¶

Fill missing values with predicted values output by a random forest regressor.

Parameters

datapandas.core.frame.DataFrame

Observed data.

columnstr

Name of the column to fill. This column should be a numeric.

targetstr, optional

Name of target column to predict. Defaults to None if data contains no target.

n_digitsint, optional

Precision in decimal digits to round the predicted values.

kwargsKeyword arguments

All keyword arguments that can be passed to sklearn.ensemble.RandomForestRegressor.

Returns

pandas.core.frame.DataFrame

Returns a data frame with missing values filled in column column.

Raises

TypeError

Raises a TypeError when column is not numeric.