[Chicago] Python 101 notebook on try.jupyter.org

Joshua Herman zitterbewegung at gmail.com
Sat Sep 5 19:57:31 CEST 2015


Dear Tanya,
I'm currently teaching myself machine learning. At first I am using
mathematica. Now I am learning more about python and setting up scikit
learn. I could probably give a 15-30 minute presentation on how to set this
up on a laptop and run the basic tests so that you can use Jupiter with
this. I haven't gotten as far as setting up theano but I think that would
be better for another workshop / presentation.
Sincerely,
Joshua Herman

On Fri, Sep 4, 2015 at 9:25 PM Tanya Schlusser <tanya at tickel.net> wrote:

> It's pretty cool. I projected it at python office hours. It's written
>> following from discussions in a paper on lessons learned from teaching
>> python to geophysics students I think? I may be mis-remembering that part.
>>
>
> You're rememering right -- this is the paper
> <http://arxiv.org/abs/1505.05425> and the tl;dr is that these geology
> professors literally guinea pigged their students for 5 years trying out
> different teaching styles (lecture + lab | lecture in lab | self-taught via
> MOOC | etc...) and mesured student satisfaction, instructor effort, and
> actual learning.
>
> The self-taught sucked, with 50% comprehension and students reviewing the
> course badly. All other forms had about 70% comprehension with varying
> levels of teacher effort and student satisfaction. The IPython notebook
> execution in-class + lecture got the highest student ratings and were the
> least instructor effort -- a major win.
>
> The try.jupyter.org project is open source
> <https://github.com/jupyter/try.jupyter.org>, so you too can host your
> own notebook server (launched using Docker images) or you can make your own
> community. This is the repo for adding community notebooks
> <https://github.com/jupyter/docker-demo-images>.
>
>
> Also -- O'Reilly may be embracing them. Here is a lecture Support Vector
> Machines given @ Pycon 2015 made in collaboration with Jake Vanderplas and
> O'Reilly:
> https://beta.oreilly.com/learning/intro-to-svm
>
> <https://beta.oreilly.com/learning/intro-to-svm>
> This is the future of education (and possibly dashboarding?)  woo-hoo!
>
> Another alternative is skulpt <http://www.skulpt.org/>...a pure
> javascript implementation of (a large chunk of) Python 2...here's the link
> to an online Python textbook with executable code using Skulpt:
> http://interactivepython.org/runestone/static/thinkcspy/index.html
>
> <http://interactivepython.org/runestone/static/thinkcspy/index.html>
>
> <http://interactivepython.org/runestone/static/thinkcspy/index.html>
> - <http://interactivepython.org/runestone/static/thinkcspy/index.html>--
> from Josh ---
>
>> Actually now that i remember it is in a notebook. Google hosts one on
>> github.
>> https://github.com/google/deepdream
>> I think we should do like a pyladies event about this if we can get it
>> working.
>>
>
>
> That would be awesome!!! looking into it...
> _______________________________________________
> Chicago mailing list
> Chicago at python.org
> https://mail.python.org/mailman/listinfo/chicago
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/chicago/attachments/20150905/851c37b8/attachment.html>


More information about the Chicago mailing list