Fix Normalize(<signed integer array>). #6837

Merged
merged 2 commits into from Aug 1, 2016

Conversation

Projects
None yet
4 participants
Contributor

anntzer commented Jul 26, 2016 edited

Basically, everything needs to be cast to a float early enough.

See #6825, #6700 -- I've decided to use the whack-a-bug-at-a-time approach for now.

@anntzer anntzer Fix Normalize(<signed integer array>).
Basically, everything needs to be cast to a float early enough.
1a2c4b0

mdboom added the needs_review label Jul 26, 2016

Owner

efiring commented Jul 26, 2016

Related to, but not part of this PR: process_value is not doing what it claims in the docstring--it is casting float32 to float64 instead of preserving it. Only 1-byte and 2-byte integers get cast to float32.

Contributor

anntzer commented Jul 27, 2016

Actually that was also broken by #6700. I fixed it now.

@anntzer anntzer Fix float normalization too (float32->float32).
ac0c85a

@tacaswell tacaswell merged commit 84d979b into matplotlib:master Aug 1, 2016

3 checks passed

continuous-integration/appveyor/pr AppVeyor build succeeded
Details
continuous-integration/travis-ci/pr The Travis CI build passed
Details
coverage/coveralls Coverage increased (+0.002%) to 70.35%
Details

tacaswell removed the needs_review label Aug 1, 2016

@tacaswell tacaswell added a commit that referenced this pull request Aug 1, 2016

@tacaswell tacaswell Merge pull request #6837 from anntzer/fix-integer-normalization
FIX: Normalize(<signed integer array>).

closes #6825
7e938fa
Owner

tacaswell commented Aug 1, 2016

backported to v2.x as 7e938fa

anntzer deleted the anntzer:fix-integer-normalization branch Aug 24, 2016

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment