Projects

This is a small list of data analysis and other related projects I have done during the last years. The code for most, if not all of them, can be found on my github account.

Data analysis projects

From time to time I publish some data analysis for fun in this blog. The code for all of them is centralized in this github repository.

Twitter deletion scripts

I quit Twitter and Facebook at the end of 2016. Before that, I coded a set of scripts to delete all old tweets and keep only the last N. If you find them useful, please use them.

WishToKnowMore

A reddit bot that summarizes links in the comment of selected posts. Currently decommissioned.

rank-es

Code here.

Takes links from the front page / pending page from meneame.net (a digg/reddit clone in Spanish) and re-ranks them using the likes and shares scores from Facebook and Twitter. Effectively, it builds a news aggregator site without users. It ran on http://rank-es.appspot.com until the Spanish government passed a stupid Intellectual Property law (more on this here, in Spanish) and I took it down.

reranker

Code here, site live here.

A better rank-es version, without Spanish media sources and with a commenting system. rank-es used to run on Google App Engine, but this is now a stand-alone application that generates static HTML files, so the whole system is quite fast.

I wrote a longer blog post about it.

Update: after some Facebook and Twitter API changes, this doesn't work anymore using guest access.

kiva_secular

Code here.

Filters Kiva's field partners according to their secular and social scores computed by this team. This is a very old project with a very basic structure, but was one of the first things I did when learning Python. Results refreshed hourly at http://www.eurielec.etsit.upm.es/~chema/kiva_secular when the university servers are up.

Kaggle

I have published the code for some Kaggle competitions I have taken part in. Here is the link to my public Kaggle profile.