Handle non-string column names #255
Conversation
|
Removing WIP. Note that the main check
is one of many ways we might want to check for strings in both Python 2 and 3. @rwedge has suggested that we have used elsewhere in Featuretools to ensure that we're consistently encoding things. This solution used fewer lines of code than a try/except block, but I don't have a strong preference for the way we should be doing this. |
Seth-Rothschild
changed the title
Seth-Rothschild
force-pushed
the
str-column-names
branch
from
4d69008
to
cc846de
Compare
featuretools/entityset/entityset.py
Outdated
| @@ -1211,6 +1211,9 @@ def _import_from_dataframe(self, | |||
| r.child_entity.id == entity_id] | |||
|
|
|||
| for c in dataframe.columns: | |||
| if not is_string(c): | |||
| raise ValueError("All column names must be strings. (Column has name {})".format(c)) | |||
There was a problem hiding this comment.
Change language to "All column names must be strings (Column {} is not a string)"
Codecov Report
@@ Coverage Diff @@
## master #255 +/- ##
==========================================
+ Coverage 94.15% 94.15% +<.01%
==========================================
Files 71 71
Lines 7644 7652 +8
==========================================
+ Hits 7197 7205 +8
Misses 447 447
Continue to review full report at Codecov.
|
|
Looks good. Merging |
close #254
This is an implementation of option 2 from #254 where we raise an error whenever we see a non
strcolumn name. There is also unit test with an integer column name to see that the error is raised.Since this is one of a few possible solutions, it's marked as WIP until we determine which path we want to follow.