from pyspark.ml.feature import StringIndexer
from pyspark.ml.feature import OneHotEncoder

# ...

def one_hot_encode(_df, input_column, output_column):
    indexer = StringIndexer(inputCol=input_column, outputCol=input_column+"_indexed", handleInvalid='skip')
    _model = indexer.fit(_df)
    _td = _model.transform(_df)
    encoder = OneHotEncoder(inputCol=input_column+"_indexed", outputCol=output_column, dropLast=True)
    _df2 = encoder.transform(_td)
    return _df2