BUG: Fix for .str.replace with invalid input #13460
Conversation
Current coverage is 84.28%@@ master #13460 diff @@ ========================================== Files 138 138 Lines 50937 50939 +2 Methods 0 0 Messages 0 0 Branches 0 0 ========================================== + Hits 42934 42936 +2 Misses 8003 8003 Partials 0 0
|
doc/source/whatsnew/v0.19.0.txt Outdated
There was a problem hiding this comment.
Pls describe the kind of error, like "does not raise ValueError".
sinhrks added Error Reporting 
priyankjain force-pushed the 13438 branch 3 times, most recently from 0e398b6 to f4f04e3 Compare 
pandas/core/strings.py Outdated
There was a problem hiding this comment.
Can you put this in the beginning? I (so switch the if/else part, also in that case the else is not needed)
IMO that is an easier to follow logic.
There was a problem hiding this comment.
Further, I don't find the error message clear. I know this is the one from the stdlib, but a numeric value can be converted to string, so that sounds a bit strange. Maybe just something like "expected a string" or "the replacement argument should be a string"
There was a problem hiding this comment.
Will do said changes, do you know why travis test has not started, it's yellow for around half an hour, no progress.
There was a problem hiding this comment.
That is because there is a queue, and each time you push a new version (or somebody else pushes on his/her PR), one travis build is added to this queue: https://travis-ci.org/pydata/pandas/pull_requests
(but will stop some of your builds)
priyankjain force-pushed the 13438 branch 2 times, most recently from 781b959 to 4079333 Compare pandas/core/strings.py Outdated
There was a problem hiding this comment.
how about "repl must be a string", to be consistent with docstring.
.str.replace now raises TypeError when replacement string is not string like
| thanks @priyankjain |
5 participants
git diff upstream/master | flake8 --diff.str.replace now raises TypeError when replacement is invalid