[Chicago] Data export ?

Lukasz Szybalski szybalski at gmail.com
Fri Sep 4 20:11:38 CEST 2015

On Fri, Sep 4, 2015 at 10:16 AM, sheila miguez <shekay at pobox.com> wrote:

> Are you thinking of providing a web site for this? I will consider the
> case where you might want to use django.
> On Thu, Sep 3, 2015 at 10:00 PM, Lukasz Szybalski <szybalski at gmail.com>
> wrote:
>> 1. Being able to connect to some database source using sqlalchemy or
>> directly.
> If this is not a flexible requirement it might be a deal breaker depending
> on your goal.
> If you use django, maybe try django-rest-pandas.
> https://github.com/wq/django-rest-pandas
> 2. Being able to display to column fields and ask the user to rearrange
>> them, select unselect them , possibly concatenate two fields,maybe
>> transform date to other format, save layout.
> I don't know d3js or other front-end stuff to know how easy it would be to
> do this. Presumably you could write something to do the above and then call
> your restful api to submit a patch or post a new thing? <hand waves>
>> 3. Then allow export to csv or xls
> built in to django-rest-pandas.
>> 4. Bonus feature: being able to schedule that export
>> 5. Bonus feature: being able to queue exports
> Doable with a cron job or something more sophisticated depending on your
> needs. For a cron job, you could set up a script to call your service api.
> For something more sophisticated you could use celery to create tasks in
> your django project.

I'm kind of looking for a finished production that would look like this:

The generation of the export table or export csv and manipulating fields is
the core what I need. While the reference to django rest pandas almost does
it; its missing the part where I change the order of fields on the fly or
in web style fashion (or plug into it via some ajax way). Good reference on
the djanog rest pandas do;

   1. Generate a CSV file via DataFrame.to_csv() (Django REST Pandas does
   this automatically)
   2. Load the file over the internet with wq/pandas.js
   3. Visualize the resulting data, perhaps with d3.js and/or wq/chart.js

I load some html UI interface. like http://www.skulpt.org/ .where you can
code python(pandas) in a browser...and press save on the final version.
Then elsewhere user presses RUN:

Phil Robare suggested:

This would give user full control over data export that would then be sent
to .

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/chicago/attachments/20150904/a11fb998/attachment.html>

More information about the Chicago mailing list