ValueError: 'multi_class' must be 'ovr' or 'multinomial'

ValueError: logistic multi_class invalid

$ python -c "from sklearn.linear_model import LogisticRegression; LogisticRegression(multi_class='one-vs-one').fit([[0],[1]],[0,1])"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
ValueError: multi_class must be 'ovr' or 'multinomial'

Why this happens

Unsupported option for multi_class.

Fix

Use ‘ovr’ or ‘multinomial’ consistent with solver.

Wrong code

from sklearn.linear_model import LogisticRegression
LogisticRegression(multi_class='one-vs-one')

Fixed code

from sklearn.linear_model import LogisticRegression
LogisticRegression(multi_class='multinomial', solver='lbfgs')