SQL connection improvements #1662

Closed
changhiskhan opened this Issue Jul 23, 2012 · 13 comments

Comments

Projects
None yet
5 participants
Contributor

changhiskhan commented Jul 23, 2012

From pydata mailing list:

Been enjoying pandas, and using it some with SQL. I've posted a gist at git://gist.github.com/3164289.git with code that sniffs DataFrame types, creates a table if needed, and writes the dataframe to Sqlite, MySQL, Oracle, or PostgreSQL.

It is lightly tested -- used on a couple of ad hoc jobs so far -- and does not save or restore indexes, but hope you'll find it useful.

changhiskhan was assigned Jul 23, 2012

Hi Chang,
Tried earlier this year to get set up for a pull request, but unfortunately had little success -- wasn't able to get the Cython to build on my 64-bit Windows box, and was therefore unable to meet the requirement to pass the set of nose tests.

I may try a Heroku set-up at some point, and that would presumably make things easier. Or if someone else is willing to integrate to to io module, they are more than welcome.

regards,
Keith

Contributor

changhiskhan commented Jul 23, 2012

No worries, I will gladly incorporate it myself. Did you have trouble building the pandas extensions or installing Cython itself?

On Jul 23, 2012, at 5:22 PM, catawbasam wrote:

Hi Chang,
Tried earlier this year to get set up for a pull request, but unfortunately had little success -- wasn't able to get the Cython to build on my 64-bit Windows box, and was therefore unable to meet the requirement to pass the set of nose tests.

I may try a Heroku set-up at some point, and that would presumably make things easier. Or if someone else is willing to integrate to to io module, they are more than welcome.

regards,
Keith


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)

Chang,

Sounds good. My problem was getting Cython installed correctly. With the 'hello world' example at http://docs.cython.org/src/userguide/tutorial.html, for example, I get:

c:\keithc\python>python setup.py build_ext --inplace
running build_ext
cythoning hello_world.pyx to hello_world.c
building 'hello_world' extension
error: Unable to find vcvarsall.bat

cheers,
Keith

-----Original Message-----
From: Chang She [mailto:reply@reply.github.com]
Sent: Monday, July 23, 2012 5:45 PM
To: Campbell, Keith C.
Subject: Re: [pandas] SQL connection improvements (#1662)

No worries, I will gladly incorporate it myself. Did you have trouble building the pandas extensions or installing Cython itself?

On Jul 23, 2012, at 5:22 PM, catawbasam wrote:

Hi Chang,
Tried earlier this year to get set up for a pull request, but unfortunately had little success -- wasn't able to get the Cython to build on my 64-bit Windows box, and was therefore unable to meet the requirement to pass the set of nose tests.

I may try a Heroku set-up at some point, and that would presumably make things easier. Or if someone else is willing to integrate to to io module, they are more than welcome.

regards,
Keith


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)

Contributor

changhiskhan commented Jul 24, 2012

You might be running into a known issue here with Cython installation on Windows. Have you looked at http://wiki.cython.org/InstallingOnWindows?

Yeah, tried installing from Gohlke. Tried Microsoft Visual Studio 2008 Express edition a couple of times with no luck. And MinGW is no good for 64-bit Win.

It would be nice if there were a way to run pandas nosetests without having to rebuild the cython stuff for changes to pure Python.

Keith

-----Original Message-----
From: Chang She [mailto:reply@reply.github.com]
Sent: Tuesday, July 24, 2012 9:34 AM
To: Campbell, Keith C.
Subject: Re: [pandas] SQL connection improvements (#1662)

You might be running into a known issue here with Cython. Have you looked at http://wiki.cython.org/InstallingOnWindows?


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)

Contributor

changhiskhan commented Jul 24, 2012

Yikes. Time for Linux ;) j/k

I agree completely on the last point. There's no good solution for now unfortunately but we'll keep it in mind and try to look for something down the road

On Jul 24, 2012, at 9:36 AM, catawbasam wrote:

Yeah, tried installing from Gohlke. Tried Microsoft Visual Studio 2008 Express edition a couple of times with no luck. And MinGW is no good for 64-bit Win.

It would be nice if there were a way to run pandas nosetests without having to rebuild the cython stuff for changes to pure Python.

Keith

-----Original Message-----
From: Chang She [mailto:reply@reply.github.com]
Sent: Tuesday, July 24, 2012 9:34 AM
To: Campbell, Keith C.
Subject: Re: [pandas] SQL connection improvements (#1662)

You might be running into a known issue here with Cython. Have you looked at http://wiki.cython.org/InstallingOnWindows?


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)

Hey Chang,

Thanks! Looks like there may be more details there than last go-around, so I'll plan on giving it another shot.

Cheers,
Keith

-----Original Message-----
From: Chang She [mailto:reply@reply.github.com]
Sent: Tuesday, July 24, 2012 6:24 PM
To: Campbell, Keith C.
Subject: Re: [pandas] SQL connection improvements (#1662)

Have you tried this: http://wiki.cython.org/64BitCythonExtensionsOnWindows


Reply to this email directly or view it on GitHub:
pydata#1662 (comment)

Contributor

danielballan commented Dec 10, 2012

Was this Keith's code ever incorporated? I wrote basic mysql functionality (https://github.com/danielballan/pandas/tree/mysql) but his cointribution more comprehensive.

Contributor

changhiskhan commented Dec 10, 2012

No they have not been incorporated. Would really appreciate the help if you can make it into a PR (yours or keith's code). We're really shorthanded here :)

Contributor

danielballan commented Dec 10, 2012

Submitted. This is my first pull request on any project, so let me know if I'm doing it wrong.

As an update to this ticket, pydata#2482 is the URL of the pull request.

Contributor

hayd commented Jul 8, 2013

Close in favour of 4163 (will migrate the 0.13 deadline ;) ).

hayd closed this Jul 8, 2013

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