reVeal.dataframe.dataframe_split

dataframe_split(df, n_parts)

Splits dataframe into the specified number of roughly equal parts. This function is similar to numpy.array_split, with a few major differences:

1. It expects an input dataframe, not an array.

2. It only accepts as input the number of input parts (am array of indices cannot be passed)

3. It does not use .swapaxes, which is deprecated and raises a FutureWarning.

4. The number of parts returned will be capped at the size of the input dataframe, whereas numpy.array_split always returns exactly n_parts, even if some parts are empty.

5. It returns a generator, not a list.

Parameters:

• df (pandas.DataFrame) – Input dataframe to split

• n_parts (int) – Number of desired parts

Yields:

pandas.DataFrame – Yields parts of split dataframe.

Raises:

ValueError – A ValueError will be raised if the numer of input sections is < 1.