This site is an archive; learn more about 8 years of OpenHatch.

[pydata-outreach-staff] Laptop setup guide, and questions

Asheesh Laroia asheesh at asheesh.org
Tue Dec 11 23:17:48 UTC 2012


On Tue, 11 Dec 2012, Chang She wrote:

> replies inline

Likewise! Thanks for the speedy reply!

> On Dec 11, 2012, at 5:24 PM, Asheesh Laroia <asheesh at asheesh.org> wrote:
>
>> Hi Chang! Thanks for replying.
>>
>> I wrote a draft of the Windows setup guide for the Pandas tutorial, which I placed here, before I saw your mail:
>>
>> https://github.com/paulproteus/PyData-Workshop-Sprint/wiki/windows-install-pandas
>>
>> If people already have Python installed, presumably they'll not be thrilled about installing a separate Python distribution from Enthought/Anaconda. I can see if it's easy to work both into the guide. I'll mull that over, and maybe try the installers and see how they work in conjunction with a Python installed from python.org.
>>
>
> There are windows binaries of most of packages out there so it's more just a matter of going through and downloading/installing each one. For mac, some packages have eggs or dmg's but many will be built from source.
>
>> You mentioned they're both free of cost; is http://www.enthought.com/products/epd_free.php the best link for that?
>
> yes, also: http://continuum.io/downloads.html
>
> EPD also gives academic access for EPD Full

Sweet. I'll look into those!

>>
>> Will your talk use ipython, and do you have bits that you'll ask people to try on their own machines? If so, then yeah, ipython for the morning does seem like a priority.
>
> Yes, I will be using IPython notebook mostly to do the talk. Probably 
> will have people try stuff on their own machines.

*nod* !

>> Similarly, can we skip the compilers stuff until the afternoon -- is 
>> the latest released pandas (for which they can install binaries) good 
>> enough for the tutorial?
>
> Hmmm…I was planning to get people started developing on pandas stuff in 
> the guided group work part of the morning session.

Exciting! In that case, it seems to me, we should front-load everything, 
rather than stick to my old separated-out plan.

Then in my opinion we should dedicate 20-30 minutes at least, in the 
morning, for laptop setup, so people can start your lecture on solid 
ground. (We'll have to do some testing and optimization of the setup 
instructions so that people can actually get through it in that much time, 
but we can do that.)

Would that be okay by you + other staffers? I'm happy to hear alternate 
proposals also.

-- Asheesh.


More information about the pydata-outreach-staff mailing list