An algorithm which can be fit on a data frame to produce a Transformer. For example, a learning algorithm is an Estimator which trains on a data frame and produces a model. It has a .fit() method because it learns (or 'fits') parameters from your data frame. The correct code snippet is:

from pyspark.ml.feature import Imputer
imputer = Imputer(strategy='median', inputCols=impute_cols, outputCols=impute_cols)
imputer_model = imputer.fit(doubles_df)
imputed_df = imputer_model.transform(doubles_df)