Behavior of regression metrics when multioutput=None is inconsistent with scikit-learn

[jameslamb]

What happened:

Using multioutput=None with the regression metrics in dask_ml.metrics.regression results in an error.

What you expected to happen:

I expected the behavior to be the same as the equivalent scikit-learn metrics functions, where multioutput=None means "all elements have the same weight".

Minimal Complete Verifiable Example:

In scikit-learn, the value of multioutput is passed through to np.average().

https://github.com/scikit-learn/scikit-learn/blob/877c6e6db42006445ccf0695c0dde3294ff4dd4a/sklearn/metrics/_regression.py#L195

np.average() treats the value None as "equally-weighted", which is the same as passing multioutput = "uniform_average". From https://numpy.org/doc/stable/reference/generated/numpy.average.html

If weights=None, then all data in a are assumed to have a weight equal to one.

import numpy as np
from sklearn.metrics import mean_squared_error

a = np.random.uniform(size=(100, 3))
b = np.random.uniform(size=(100, 3))

raw_output = mean_squared_error(a, b, multioutput="raw_values")
print(raw_output)

# [0.17118814 0.15964742 0.13095381]

mean_squared_error(a, b, multioutput=None)

# 0.15392978995168105

In dask-ml, passing multioutput=None results in an error.

import dask.array as da
from dask_ml.metrics.regression import mean_squared_error

a = da.random.uniform(size=(100, 3))
b = da.random.uniform(size=(100, 3))

raw_output = mean_squared_error(a, b, multioutput="raw_values")
print(raw_output)
# dask.array<mean_agg-aggregate, shape=(3,), dtype=float64, chunksize=(3,), chunktype=numpy.ndarray>

mean_squared_error(a, b, multioutput=None)
# ValueError: Weighted 'multioutput' not supported.

Anything else we need to know?:

Environment:

• Dask version:

output of 'conda list | grep -E "dask|distributed"'

dask                      2021.3.0           pyhd3eb1b0_0  
dask-core                 2021.3.0           pyhd3eb1b0_0  
dask-glm                  0.2.0                    pypi_0    pypi
dask-saturn               0.2.2                    pypi_0    pypi
dask-sphinx-theme         1.3.5                    pypi_0    pypi
distributed               2021.3.0         py37h06a4308_0  

• Python version, Operating System

output of 'conda info'

     active environment : None
       user config file : /home/jlamb/.condarc
 populated config files : 
          conda version : 4.9.2
    conda-build version : not installed
         python version : 3.7.6.final.0
       virtual packages : __glibc=2.27=0
                          __unix=0=0
                          __archspec=1=x86_64
       base environment : /home/jlamb/miniconda3  (writable)
           channel URLs : https://repo.anaconda.com/pkgs/main/linux-64
                          https://repo.anaconda.com/pkgs/main/noarch
                          https://repo.anaconda.com/pkgs/r/linux-64
                          https://repo.anaconda.com/pkgs/r/noarch
          package cache : /home/jlamb/miniconda3/pkgs
                          /home/jlamb/.conda/pkgs
       envs directories : /home/jlamb/miniconda3/envs
                          /home/jlamb/.conda/envs
               platform : linux-64
             user-agent : conda/4.9.2 requests/2.22.0 CPython/3.7.6 Linux/5.4.0-70-generic ubuntu/18.04.4 glibc/2.27
                UID:GID : 1000:1000
             netrc file : None
           offline mode : False

• Install method (conda, pip, source): installed from source as of f5e5bb4, using python setup.py install.