Fix for GH 6885 - get_dummies chokes on unicode values #6975
Conversation
|
Note that the failed test wasn't related to my changes: |
|
can you add a release note in Bug Fixes referecing this issue, otherwise looks good @hayd @TomAugspurger ok by you? |
| eacute = unicodedata.lookup('LATIN SMALL LETTER E WITH ACUTE') | ||
| s = [e, eacute, eacute] | ||
| res = get_dummies(s, prefix='letter') | ||
| print(res) |
There was a problem hiding this comment.
Was this print in here just for debugging? It's not actually checking anything right? You should remove it if so.
|
The travis failure looked like an unrelated yahoo test. I restarted it on Travis (I think that's a new feature). We'll see if Github notices and updates with the results. |
|
Apologies for the stray print - I'll remove that and then add the edit in bugfixes |
|
@maxgrenderjones did you force push? (your old commit is still here) |
| @@ -270,6 +270,7 @@ Improvements to existing features | |||
| Bug Fixes | |||
| ~~~~~~~~~ | |||
|
|
|||
| - Bug causing UnicodeEncodeError when get_dummies called with unicode values and a prefix (:issue:`6885`) | |||
There was a problem hiding this comment.
usually we put the new fixes at the end, but this is ok
|
ping me when green |
|
Last code-changing build is green - don't know if travis bothers building if you just change docs. Apologies that a 4 character change took four commits! |
|
merged via 7918d60 thanks! |
closes #6885
Please be gentle - this is the first time I've tried to contribute either to someone else's python project or another project on github, so if my python or git foo is lacking, I apologise in advance!