These are chat archives for FreeCodeCamp/DataScience

25th
Jan 2016
@evaristoc d3 is only really useful if you really need that level of control. if you're creating basic scatterplots or charts available as presets, there are some wrappers for d3 you can use
there are many
or you could just use a more conventional charting library
i've used google charts. it has some nice basic data table functions too
i would like to see a library that can render PNGs so the charts could be displayed also here in chat
this looks cool if you're on react too
evaristoc
@evaristoc
Jan 25 2016 10:42 UTC

@dcsan thanks... yes: I would definitively have a look at those options...

If I am emphasising d3 is because I believe that I will learn more about event-driven programming, css/html and JS as well as getting better in interactive visualization techniques all at once if I follow that path, but definitively: no point to use d3 if what you need is just a traditional graph...

CamperBot
@camperbot
Jan 25 2016 10:42 UTC
evaristoc sends brownie points to @dcsan :sparkles: :thumbsup: :sparkles:
:star: 143 | @dcsan | http://www.freecodecamp.com/dcsan
evaristoc
@evaristoc
Jan 25 2016 11:01 UTC

Hi @/all

THE DSR WEEKLY DIGEST:

DSR (DataScience Room) is an effort to gather campers of all levels and specializations willing to engage in discussion, collaboration and practice of data-related projects with focus on JS and FCC.

Previous Digest

Hi EVERYONE!

Question: who have been playing with the dataset recently? Talk about your results here!


Last Week...:


Some Proposals:
--- @QuincyLarson suggested that visualization and analyses contributions could be included in the FCC site, so feel free to participate!
--- check also the @vicky002's list


In the FCC News...:

  • I mentioned several reviews about visualization and d3: I decided to include a link to the one discussing when d3 should be used: "You may not need d3"

Please also check the @dcsan list of chart libraries


... Next Week...:

  • focusing on the projects already discussed in previous weeks

Wish List:

  • creating small teams to analyse different aspects of the data and come every end of the month, for example, in order to put all our findings together for FCC to check
evaristoc
@evaristoc
Jan 25 2016 22:56 UTC

for those interested on machine learning to study some interesting patterns, application for MOOC:

http://web.stanford.edu/~cpiech/bio/papers/Uncovering%20Patterns%20in%20Student%20Work.pdf
@quatridunull, @vicky002 --- let you know that we don't have enough data for that in the Torrent dataset, BUT we can start something with Gitter rooms: https://www2.informatik.hu-berlin.de/~kloftmar/publications/emnlp_mooc.pdf

This one was what I was looking for actually (the rest was found meanwhile but still interesting...), @luishendrix92 --- don't get scare, but we can at least get inspired from this work, also @hallaathrad, @ltegman:
http://web.stanford.edu/~cpiech/bio/papers/homeworkVariance.pdf


Differences between ML and DataSci:
https://hackbrightacademy.com/blog/data-science-machine-learning/


@jameswinegar,@jbmartinez: this one interesting but maybe still a bit too much... ML extension to peer reviewing:
http://hci.stanford.edu/publications/2015/Flock/flock_paper.pdf