Skip to content

Pandas 0.14 broke Python 2.6 support #7284

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
jd opened this issue May 30, 2014 · 25 comments · Fixed by #7285
Closed

Pandas 0.14 broke Python 2.6 support #7284

jd opened this issue May 30, 2014 · 25 comments · Fixed by #7285
Labels
Bug Compat pandas objects compatability with Numpy or Python functions IO Excel read_excel, to_excel
Milestone

Comments

@jd
Copy link
Contributor

jd commented May 30, 2014

Since today I've the following issue when importing Pandas on Python 2.6

2014-05-30 15:14:29.999 |   File "gnocchi/carbonara.py", line 22, in <module>
2014-05-30 15:14:29.999 |     import pandas
2014-05-30 15:14:29.999 |   File "/home/jenkins/workspace/gate-gnocchi-python26/.tox/py26/lib/python2.6/site-packages/pandas/__init__.py", line 45, in <module>
2014-05-30 15:14:30.000 |     from pandas.io.api import *
2014-05-30 15:14:30.000 |   File "/home/jenkins/workspace/gate-gnocchi-python26/.tox/py26/lib/python2.6/site-packages/pandas/io/api.py", line 7, in <module>
2014-05-30 15:14:30.000 |     from pandas.io.excel import ExcelFile, ExcelWriter, read_excel
2014-05-30 15:14:30.000 |   File "/home/jenkins/workspace/gate-gnocchi-python26/.tox/py26/lib/python2.6/site-packages/pandas/io/excel.py", line 626, in <module>
2014-05-30 15:14:30.000 |     .format(openpyxl_compat.start_ver, openpyxl_compat.stop_ver))
2014-05-30 15:14:30.000 | ValueError: zero length field name in format
@jreback
Copy link
Contributor

jreback commented May 30, 2014

hmm, very odd, can you show pd.show_versions()

@cpcloud
Copy link
Member

cpcloud commented May 30, 2014

i got this i see the problem

@jreback
Copy link
Contributor

jreback commented May 30, 2014

gr8!

@jreback
Copy link
Contributor

jreback commented May 30, 2014

hmm..very odd as travis passes just fine with an older version of openpyxl

@jreback jreback added this to the 0.14.1 milestone May 30, 2014
@jreback
Copy link
Contributor

jreback commented May 30, 2014

cc @neirbowj

@cpcloud
Copy link
Member

cpcloud commented May 30, 2014

it's because it's just a warning and maybe tests don't cover on the one it runs

@cpcloud
Copy link
Member

cpcloud commented May 30, 2014

actually all versions running on travis are >= 1.6.2 so the warning won't be triggered.. should i put on 2.6 req file?

@jreback
Copy link
Contributor

jreback commented May 30, 2014

yes, change that to 2.0.3 so it will trigger the warning (and fail)

@cpcloud
Copy link
Member

cpcloud commented May 30, 2014

cool

@jreback
Copy link
Contributor

jreback commented May 30, 2014

@jd in the mean time if you downgrade openpyxl to < version 2 this would work I believe
>= 2.0 in openpyxl is backward incompatible so in 0.14.0 we provide this warning (which unfortunately was not tested with the newest version on 2.6), oversite.

@colindickson
Copy link

i don't have - nor do i want - openpyxl and i get the warning when i import pandas. i commented out the warning in my copy of the code, but i hope the warning will be completely removed in 0.14.1.

@jd
Copy link
Contributor Author

jd commented Jun 2, 2014

@jreback I don't actually have openpyxl at all. Any ETA on the fix/release? It's clearly blocking our project :(

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

well easiest to simply install openpyxl < 2 and it will work. master will be fixed soon as well. 0.14.1 is slated for about 1 month.

@jorisvandenbossche
Copy link
Member

@jreback Would this not be a reason to do a quicker bugfix release? As this seems rather critical to me (pandas broken for everybody who still uses 2.6 (unless you have openpyxl intalled, but if you don't use read_excel this is not that likely).

We can always still do a 0.14.2 release in a month or so (but I don't know how much work it is do to this)

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

since installing openpyxl fixes this I don't think it's a reason to do a point release and/updating to master

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

furthermore op can simply downgrade to 0.13.1

@jorisvandenbossche
Copy link
Member

OK, but it is not really clear from the message that that is the solution, you already have to find this discussion to know that.

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

it's an issue only on. 2.6 if you don't have openpyxl installed at all - if it's installed and. < 2 thrn it won't come up

so you would have to have an existing. 2.6 installation have updated openpyxl in last. couple of weeks and upgrade pandas

prob not a lot of users

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

@colindickson @jd what platforms are you on?

@jd
Copy link
Contributor Author

jd commented Jun 2, 2014

@jreback Ubuntu 12.04 LTS

@colindickson
Copy link

i have python 3.4 (x64) on both windows and ubuntu.

for the record, my only issue is with the warning message itself. i do not have openpyxl, and i don't really use much excel I/O - when i do, i use xlxswriter, since it is faster.

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

ok...on ubuntu you can build from master if you'd like, see here: http://pandas-docs.github.io/pandas-docs-travis/install.html#installing-from-source

for windows I post binaries of master here: http://pandas.pydata.org/pandas-build/dev/
which I will do tonight

@jd
Copy link
Contributor Author

jd commented Jun 2, 2014

@jreback Oh I know, but this is not an option to hack the entire OpenStack CI system to do that :) Nevermind I've forced a dependency on pandas<0.14 for now. Thanks for fixing!

@jreback
Copy link
Contributor

jreback commented Jun 2, 2014

gr8!

@cpcloud
Copy link
Member

cpcloud commented Jun 2, 2014

This was a fun one. Started out like 😎 ended up like 😡

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Bug Compat pandas objects compatability with Numpy or Python functions IO Excel read_excel, to_excel
Projects
None yet
Development

Successfully merging a pull request may close this issue.

5 participants