DOC: Changed docstring style for optional arguments

Author: MechCoder

sklearn/linear_model/logistic.py

@@ -372,7 +372,7 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
  To lessen the effect of regularization on synthetic feature weight
  (and therefore on the intercept) intercept_scaling has to be increased.
 
- multi_class : str, optional default 'ovr'
+ multi_class : str, {'ovr', 'multinomial'}
  Multiclass option can be either 'ovr' or 'multinomial'. If the option
  chosen is 'ovr', then a binary problem is fit for each label. Else
  the loss minimised is the multinomial loss fit across
@@ -455,10 +455,11 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
 
  else:
  lbin = LabelBinarizer()
- Y = lbin.fit_transform(y)
- if Y.shape[1] == 1:
- Y = np.hstack([1 - Y, Y])
- w0 = np.zeros((Y.shape[1], n_features + int(fit_intercept)), order='F')
+ Y_bin = lbin.fit_transform(y)
+ if Y_bin.shape[1] == 1:
+ Y_bin = np.hstack([1 - Y_bin, Y_bin])
+ w0 = np.zeros((Y_bin.shape[1], n_features + int(fit_intercept)),
+ order='F')
  mask_classes = classes
 
  if class_weight == "auto":
@@ -495,7 +496,7 @@ def logistic_regression_path(X, y, pos_class=None, Cs=10, fit_intercept=True,
  for C in Cs:
  if solver == 'lbfgs':
  if multi_class == 'multinomial':
- target = Y
+ target = Y_bin
  func = _multinomial_loss_grad
  else:
  target = y
@@ -627,7 +628,7 @@ def _log_reg_scoring_path(X, y, train, test, pos_class=None, Cs=10,
  To lessen the effect of regularization on synthetic feature weight
  (and therefore on the intercept) intercept_scaling has to be increased.
 
- multi_class : str, optional default 'ovr'
+ multi_class : str, {'ovr', 'multinomial'}
  Multiclass option can be either 'ovr' or 'multinomial'. If the option
  chosen is 'ovr', then a binary problem is fit for each label. Else
  the loss minimised is the multinomial loss fit across
@@ -658,8 +659,10 @@ def _log_reg_scoring_path(X, y, train, test, pos_class=None, Cs=10,
  log_reg._enc = LabelEncoder()
  if multi_class == 'ovr':
  log_reg._enc.fit([-1, 1])
- else:
+ elif multi_class == 'multinomial':
  log_reg._enc.fit(np.unique(y_train))
+ else:
+ raise ValueError("multi_class should be either multinomial or ovr")
 
  if pos_class is not None:
  mask = (y_test == pos_class)
@@ -772,7 +775,7 @@ class LogisticRegression(BaseLibLinear, LinearClassifierMixin,
  tol : float, optional
  Tolerance for stopping criteria.
 
- multi_class : str, optional default 'ovr'
+ multi_class : str, {'ovr', 'multinomial'}
  Multiclass option can be either 'ovr' or 'multinomial'. If the option
  chosen is 'ovr', then a binary problem is fit for each label. Else
  the loss minimised is the multinomial loss fit across
@@ -958,7 +961,7 @@ class LogisticRegressionCV(LogisticRegression, BaseEstimator,
  Otherwise the coefs, intercepts and C that correspond to the
  best scores across folds are averaged.
 
- multi_class : str, optional default 'ovr'
+ multi_class : str, {'ovr', 'multinomial'}
  Multiclass option can be either 'ovr' or 'multinomial'. If the option
  chosen is 'ovr', then a binary problem is fit for each label. Else
  the loss minimised is the multinomial loss fit across