Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

Loading…

make sure there aren't any deprecation warnings that incorrectly say stuff will happen in 1.8 #3228

Closed
njsmith opened this Issue · 4 comments

3 participants

@njsmith
Owner

There are a number of deprecations that we expected would be finalized in 1.8, but that won't be, e.g. gh-596, gh-456, gh-294. We've bumped them to reconsider for 1.9, but we should run a grep over the source to make sure that we haven't left any behind any messages that say "Warning: blah balh will happen in numpy 1.8", and if there are any we should replace them with something more vague.

@seberg
Owner

Just did a quick grep. It seems that the only occasion that mentions either 1.8. or 1.9. together with deprecations is the diagonal documentation.

@charris
Owner

There is a test_diagonal_deprecation that checks for FutureWarning. I believe that warning was for 1.7 and we should be returning a ro view of the diagonal. Need to check that and remove the warning and test.

@charris
Owner

Here is the FutureWarning message

        "Numpy has detected that you (may be) writing to an array returned\n"
        "by numpy.diagonal or by selecting multiple fields in a record\n"
        "array. This code will likely break in the next numpy release --\n"
        "see numpy.diagonal or arrays.indexing reference docs for details.\n"
        "The quick fix is to make an explicit copy (e.g., do\n"
        "arr.diagonal().copy() or arr[['f0','f1']].copy()).";

What is the multiple fields thing?

@charris
Owner

See #3604 for the diagonal fixup.

@charris charris referenced this issue from a commit in charris/numpy
@charris charris DOC: Remove references to diagonal changes in 1.8.
Not happening. Also remove reference to selecting multiple fields out of
an array returning a view from 1.8.0 release notes.

Closes #3228.
df1aa99
@charris charris closed this in #3608
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.