/pandas

New issue

BUG: Series.align(join='left') and join='right' are inverted #10665

Closed
gdementen opened this Issue Jul 24, 2015 · 3 comments

Comments

Assignees
No one assigned
Labels
Bug Reshaping
Projects
None yet
Milestone
  0.17.0
3 participants
@gdementen @sinhrks @jreback
@gdementen
Contributor

gdementen commented Jul 24, 2015

join='left' does a right join and 'right' does a left join

See the notebook for details
https://nbviewer.jupyter.org/gist/gdementen/fa4d388993ee981b5698
@gdementen
Contributor

gdementen commented Jul 24, 2015

btw, this is with pandas 0.16.2

gdementen changed the title from Series.align(join='left') and join='right' are inverted to BUG/Series.align(join='left') and join='right' are inverted Jul 24, 2015

gdementen changed the title from BUG/Series.align(join='left') and join='right' are inverted to BUG: Series.align(join='left') and join='right' are inverted Jul 24, 2015

@gdementen
Contributor

gdementen commented Jul 24, 2015

probably related to (possibly a dupe of, I am not sure) #10147.
The strange thing, is that for DataFrames, align seems to work correctly.
@sinhrks
Member

sinhrks commented Jul 29, 2015

Thanks. Looks to be caused by the following lines.

When Series with MultiIndex is on left side, how=left and right must be flipped also. PR is appreciated!

@sinhrks sinhrks added Bug Reshaping labels Jul 29, 2015

This was referenced Aug 1, 2015

Merged
Closed

jreback added this to the 0.17.0 milestone Aug 3, 2015

jreback closed this in #10716 Aug 18, 2015

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