pyspark.sql.DataFrame.randomSplit

DataFrame.randomSplit(weights, seed=None)[source]

Randomly splits this DataFrame with the provided weights.

New in version 1.4.0.

Parameters:
weightslist

list of doubles as weights with which to split the DataFrame. Weights will be normalized if they don’t sum up to 1.0.

seedint, optional

The seed for sampling.

Examples

>>> splits = df4.randomSplit([1.0, 2.0], 24)
>>> splits[0].count()
2
>>> splits[1].count()
2