/pandas

New issue

Unexpected behavior with groupby on single-row dataframe? #11741

Closed
dwyatte opened this Issue Dec 2, 2015 · 3 comments

Comments

Assignees
No one assigned
Labels
Bug Groupby
Projects
None yet
Milestone
  0.18.0
3 participants
@dwyatte @nbonnotte @jreback
@dwyatte

dwyatte commented Dec 2, 2015

Do single-row dataframes (not series) get special treatment in some way? It seems you can do arbitrary groupby operations on them on non-existent columns without errors.

df1 = pd.DataFrame(np.random.randn(1,4), columns=list('ABCD'))
df2 = pd.DataFrame(np.random.randn(2,4), columns=list('ABCD'))

In [3]: df1.groupby('asdf')
Out[3]: pandas.core.groupby.DataFrameGroupBy object at 0x1139acb10

In [4]: df2.groupby('asdf')
KeyError: 'asdf'
@nbonnotte
Contributor

nbonnotte commented Jan 16, 2016

Yeah, there is a problem. I discovered that while working on issue #11640

I've a PR ready

nbonnotte referenced this issue Jan 16, 2016

Closed

TST in .drop and .groupby for dataframes with multi-indexed columns #11717

nbonnotte referenced this issue Jan 16, 2016

Closed

BUG in .groupby for single-row DF #12063

@nbonnotte
Contributor

nbonnotte commented Jan 16, 2016

Solved with PR #12063

jreback added the Groupby label Jan 16, 2016

jreback added this to the 0.18.0 milestone Jan 16, 2016

jreback added the Bug label Jan 16, 2016

@nbonnotte nbonnotte added a commit to nbonnotte/pandas that referenced this issue Jan 16, 2016

@nbonnotte nbonnotte BUG in .groupby for single-row DF 
No KeyError was raised when grouping by a non-existant column

Fixes #11741

Xref issue #11640, PR #11717
 a5c4555

jreback closed this in e9e8598 Jan 17, 2016

@dwyatte

dwyatte commented Jan 19, 2016

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