My name is Marcelo, I was born in Brazil and live in Berlin. I believe that my educational background as an economist with focus on causal inference, coupled with my interests in programming as well as being a long time Linux user and tinkerer put me in the right track to learn and grow in the data science field.
If you got here from my job application, please have a look around the pinned and following repositories.
- logitModel: Final project of "Statistical Programming with R" course, where I wrote a package to run a logit estimation from scratch, including an inplementation of the Newton–Raphson optimization algorithm.
- enetfactorforecastR: Final project of “Multivariate Verfahren” course, where we developed a method for variable selection and forecasting using Principal Components and Elastic Net for dimensionality reduction. In this group work I was mostly responsible for the elastic net implementation and the visualizations.
- kaleckianModel: [R Shiny] In my exchange year at the Unicamp I wrote a very simple solver of a Kaleckian model using Shiny, used for illustrative purpose in a learning context. Here you can check the interactive app, but text is in portuguese.
- SOEPutils: Helper Stata programs for wrangling SOEP data. This repository contains a series of Stata programs useful to deal with several steps involved in the generation of SOEP data.
- mat2tex: A simple package to export a stata matrix to a body of a latex table.
- global-terrorism-data: Final project of “Betriebssystem und Programmierung – Python” written with Marek Felten. A interactive result of this project can be found in here.
- acarril/StataLinux: I also made a few contributions to the StataLinux Plugin. This plugins enables the user to send code to Stata from the Sublime Text editor.
Sometimes you can also find me answering and asking some questions at StackOverflow.