Add zero_division for single class prediction in MCC #28982
Description
Describe the bug
I have found a potential edge case issue with sklearn.metrics.matthews_corrcoef. The example provided in the documentation works as expected:
from sklearn.metrics import matthews_corrcoef
matthews_corrcoef([1, 1, 1, -1], [1, -1, 1, 1]) # returns -1/3 OKHowever, edge cases appear when either y or y_true have only one single label.
Steps/Code to Reproduce
from sklearn.metrics import matthews_corrcoef
matthews_corrcoef([1, 1, 1, 1], [1, 1, 1, 1]) # returns 0 instead of 1
matthews_corrcoef([0, 0, 0, 0], [0, 0, 0, 0]) # returns 0 instead of 1Expected Results
Outputs should be 1, not 0.
Actual Results
Outputs are 0.
Versions
System:
python: 3.11.9 (main, Apr 2 2024, 08:25:04) [Clang 16.0.6 ]
executable: /tmp/sklearn-test/.venv/bin/python
machine: macOS-14.2.1-arm64-arm-64bit
Python dependencies:
sklearn: 1.4.2
pip: 24.0
setuptools: 69.5.1
numpy: 1.26.4
scipy: 1.13.0
Cython: None
pandas: None
matplotlib: None
joblib: 1.4.2
threadpoolctl: 3.5.0
Built with OpenMP: True
threadpoolctl info:
user_api: openmp
internal_api: openmp
num_threads: 8
prefix: libomp
filepath: /tmp/sklearn-test/.venv/lib/python3.11/site-packages/torch/lib/libomp.dylib
version: None
user_api: blas
internal_api: openblas
num_threads: 8
prefix: libopenblas
filepath: /tmp/sklearn-test/.venv/lib/python3.11/site-packages/numpy/.dylibs/libopenblas64_.0.dylib
version: 0.3.23.dev
threading_layer: pthreads
architecture: armv8
user_api: blas
internal_api: openblas
num_threads: 8
prefix: libopenblas
filepath: /tmp/sklearn-test/.venv/lib/python3.11/site-packages/scipy/.dylibs/libopenblas.0.dylib
version: 0.3.26.dev
threading_layer: pthreads
architecture: neoversen1
user_api: openmp
internal_api: openmp
num_threads: 8
prefix: libomp
filepath: /tmp/sklearn-test/.venv/lib/python3.11/site-packages/sklearn/.dylibs/libomp.dylib
version: None