cv=1 is meaningless: many classifiers need at least 2 training samples from 2 different classes to just know the kind of problems they are addressing and cannot be fitted on an empty trainingset.

But I agree we at least need better error messages. For instance now we have:

>>> from sklearn.svm import LinearSVC
>>> clf = LinearSVC()

>>> from sklearn.datasets import load_digits
>>> digits = load_digits()

>>> from sklearn.grid_search import GridSearchCV
>>> GridSearchCV(clf, {'C': [1, 10]}, cv=1).fit(digits.data, digits.target)
Traceback (most recent call last):
  File "<ipython-input-54-9e81f5317334>", line 1, in <module>
    GridSearchCV(clf, {'C': [1, 10]}, cv=1).fit(digits.data, digits.target)
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 700, in fit
    return self._fit(X, y, ParameterGrid(self.param_grid), **params)
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 483, in _fit
    for parameters in parameter_iterable
  File "/Users/ogrisel/code/scikit-learn/sklearn/externals/joblib/parallel.py", line 514, in __call__
    self.dispatch(function, args, kwargs)
  File "/Users/ogrisel/code/scikit-learn/sklearn/externals/joblib/parallel.py", line 311, in dispatch
    job = ImmediateApply(func, args, kwargs)
  File "/Users/ogrisel/code/scikit-learn/sklearn/externals/joblib/parallel.py", line 135, in __init__
    self.results = func(*args, **kwargs)
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 286, in fit_grid_point
    clf.fit(X_train, y_train, **fit_params)
  File "/Users/ogrisel/code/scikit-learn/sklearn/svm/base.py", line 668, in fit
    raise ValueError("The number of classes has to be greater than"
ValueError: The number of classes has to be greater than one.

>>> GridSearchCV(clf, {}, cv=2).fit(digits.data, digits.target)
Traceback (most recent call last):
  File "<ipython-input-55-0fd69621ac4a>", line 1, in <module>
    GridSearchCV(clf, {}, cv=2).fit(digits.data, digits.target)
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 700, in fit
    return self._fit(X, y, ParameterGrid(self.param_grid), **params)
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 483, in _fit
    for parameters in parameter_iterable
  File "/Users/ogrisel/code/scikit-learn/sklearn/externals/joblib/parallel.py", line 513, in __call__
    for function, args, kwargs in iterable:
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 480, in <genexpr>
    delayed(fit_grid_point)(
  File "/Users/ogrisel/code/scikit-learn/sklearn/grid_search.py", line 84, in __iter__
    keys, values = zip(*items)
ValueError: need more than 0 values to unpack

Uh oh!

allow GridSearchCV to work with params={} or cv=1 #2048

Description

Metadata

Metadata

Assignees

Labels

Type

Fields

Projects

Milestone

Relationships

Development

Issue actions