LinearDiscriminantAnalysis.fit() fails if #samples == #labels. #12374
Description
Description
The LinearDiscriminantAnalysis.fit() method throws an exception if number of samples and number of labels is the same, i.e. each label has exactly one sample.
Steps/Code to Reproduce
>>> import sklearn.discriminant_analysis
>>> import numpy
>>> X = numpy.array([[0.5, 0.6], [0.6, 0.5]])
>>> y = numpy.array(["a", "b"])
>>> lda = sklearn.discriminant_analysis.LinearDiscriminantAnalysis()
>>> lda.fit(X, y)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/…/lib/python3.6/site-packages/sklearn/discriminant_analysis.py", line 455, in fit
self._solve_svd(X, y)
File "/…/lib/python3.6/site-packages/sklearn/discriminant_analysis.py", line 379, in _solve_svd
fac = 1. / (n_samples - n_classes)
ZeroDivisionError: float division by zeroExpected Results
No exception, and a successful call to fit().
Actual Results
Exception, see above.
Versions
>>> sklearn.show_versions()
System
------
python: 3.6.6 (default, Jun 28 2018, 05:50:42) [GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.42.1)]
executable: /…/bin/python
machine: Darwin-15.6.0-x86_64-i386-64bit
BLAS
----
macros: NO_ATLAS_INFO=3, HAVE_CBLAS=None
lib_dirs:
cblas_libs: cblas
Python deps
-----------
pip: 18.0
setuptools: 28.8.0
sklearn: 0.20.0
numpy: 1.15.2
scipy: 1.1.0
Cython: None
pandas: None