rankdata#

named_arrays.stats.rankdata(a, axis=None, method='average', where=True)#

Assign ranks to the data along the given axis, dealing with ties according to the chosen method.

This is a named_arrays implementation of scipy.stats.rankdata().

Parameters:

  • a (AT) – The input array to rank.

  • axis (None | str | Sequence[str]) – The axis or axes along which to compute the ranks. If None (the default), the ranks are computed over a flattened version of all the axes.

  • method (str) –

    The convention used to assign ranks to tied values.

    • "average": The average of the ranks spanned by the tied values.

    • "min": The minimum of the ranks spanned by the tied values (also known as competition ranking).

    • "max": The maximum of the ranks spanned by the tied values.

    • "dense": Like "min", but the rank of the next distinct value is always one greater than the previous group (no gaps).

  • where (bool | AbstractScalarArray) – A boolean mask selecting which elements to include in the ranking. Excluded elements are assigned a rank of NaN.

Return type:

AT

See also

scipy.stats.rankdata()

The equivalent Scipy function.