[MRG + 1] Fix negative inputs checking in mean_squared_log_error #9968

gustavodemari · 2017-10-21T19:16:05Z

Reference Issues/PRs

What does this implement/fix? Explain your changes.

The logic below had some flaw when just one array has some negative value.

if not (y_true >= 0).all() and not (y_pred >= 0).all():
   raise ValueError("Mean Squared Logarithmic Error cannot be used when "
                         "targets contain negative values.")

For example:

import numpy as np

y_true = np.array([1, 2, 3])
y_pred = np.array([1, -2, 3])

not (y_true >= 0).all()
False

not (y_pred >= 0).all()
True

not (y_true >= 0).all() and not (y_pred >= 0).all()
False

Now, with the change this doesn't occurs anymore.

(y_true < 0).any() or (y_pred < 0).any()
True

Any other comments?

amueller · 2017-10-21T22:53:21Z

can you please add a regression test?

gustavodemari · 2017-10-21T23:03:09Z

This line doesn't already do this? @amueller

scikit-learn/sklearn/metrics/tests/test_regression.py

Lines 64 to 66 in ef5cb84

    
           assert_raises_regex(ValueError, "Mean Squared Logarithmic Error cannot be " 
        
                               "used when targets contain negative values.", 
        
                               mean_squared_log_error, [-1.], [-1.])

If not, I can add this line?

assert_raises_regex(ValueError, "Mean Squared Logarithmic Error cannot be "
                        "used when targets contain negative values.",
                        mean_squared_log_error, [1, 2, 3], [1, -2, 3])

Thanks

amueller · 2017-10-21T23:36:16Z

thanks. The previous test passed with master, which means it's not strict enough.

amueller · 2017-10-24T22:02:21Z

lgtm

jnothman · 2017-10-25T08:54:16Z

Thanks

…cs/donigian-update-contribution-guidelines * 'master' of github.com:scikit-learn/scikit-learn: (23 commits) fixes scikit-learn#10031: fix attribute name and shape in documentation (scikit-learn#10033) [MRG+1] add changelog entry for fixed and merged PR scikit-learn#10005 issue scikit-learn#9633 (scikit-learn#10025) [MRG] Fix LogisticRegression see also should include LogisticRegressionCV(scikit-learn#9995) (scikit-learn#10022) [MRG + 1] Labels of clustering should start at 0 or -1 if noise (scikit-learn#10015) MAINT Remove redundancy in scikit-learn#9552 (scikit-learn#9573) [MRG+1] correct comparison in GaussianNB for 'priors' (scikit-learn#10005) [MRG + 1] ENH add check_inverse in FunctionTransformer (scikit-learn#9399) [MRG] FIX bug in nested set_params usage (scikit-learn#9999) [MRG+1] Fix LOF and Isolation benchmarks (scikit-learn#9798) [MRG + 1] Fix negative inputs checking in mean_squared_log_error (scikit-learn#9968) DOC Fix typo (scikit-learn#9996) DOC Fix typo: x axis -> y axis (scikit-learn#9985) improve example plot_forest_iris.py (scikit-learn#9989) [MRG+1] Deprecate pooling_func unused parameter in AgglomerativeClustering (scikit-learn#9875) DOC update news DOC Fix three typos in manifold documentation (scikit-learn#9990) DOC add missing dot in docstring DOC Add what's new for 0.19.1 (scikit-learn#9983) Improve readability of outlier detection example. (scikit-learn#9973) DOC: Fixed typo (scikit-learn#9977) ...

…kit-learn#9968) * fixes msle when the inputs is negative, resolves scikit-learn#9963 * adding some regression tests for msle metric

fixes msle when the inputs is negative, resolves scikit-learn#9963

b6b9223

adding some regression tests for msle metric

7f5b189

amueller changed the title ~~[MRG] Fix negative inputs checking in mean_squared_log_error~~ [MRG + 1] Fix negative inputs checking in mean_squared_log_error Oct 24, 2017

jnothman merged commit f84581b into scikit-learn:master Oct 25, 2017

jnothman mentioned this pull request Aug 29, 2018

IsolationForest with categorical attributes #11927

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[MRG + 1] Fix negative inputs checking in mean_squared_log_error #9968

[MRG + 1] Fix negative inputs checking in mean_squared_log_error #9968

[MRG + 1] Fix negative inputs checking in mean_squared_log_error #9968

[MRG + 1] Fix negative inputs checking in mean_squared_log_error #9968

Conversation

Reference Issues/PRs

What does this implement/fix? Explain your changes.

Any other comments?