Software

7.8.2015

This is an incomplete index of software I've written. More and most recent activity can be found on my gitlab profile page and/or on github.

R package: gitlabr

gitlabr is a client for the gitlab API written in R. It's currently in early development and has only basic features (listing and reading issues, repositories, ...), but what is there, is working. You can find and install it from my gitlab instance: http://gitlab.points-of-interest.cc/points-of-interest/gitlabr

R package: manifestoR

manifestoR is an R package that provides access to coded election programmes from the Manifesto Corpus and to the Manifesto Project's Main Dataset. The Manifesto Project collects and analyses election programmes across time and space to measure the political preferences of parties. The Manifesto Corpus contains these programmes in the Corpus format of the R package 'tm' to enable easy use of text processing and text mining functionality.

manifestoR also includes functions for scaling coded political texts and infrastructure to easily build custom scales.

You can find manifestoR on CRAN or a development version on github.

Archive

Everything below is software I am not currently working on, although it is stable and working if not stated otherwise.

R package: QCAtools

QCAtools is a small R package to make it more convenient and interactive to use the R programming language for Qualitative Comparative Analysis (QCA), a qualitative, but formal research approach based on Boolean algebra. QCAtools provides functions to

  • evaluate and plot Boolean formulae on fuzzy set score data
  • apply Boolean operations to fuzzy set score data
  • compute consistency and coverage measures as used in QCA.

The idea of QCAtools is to allow the user to focus on the logic, the model and the data rather than the programming by treating written Boolean formulae (strings), R functions and data matrices with fuzzy set score values just "as you would expect". In addition, either of them can be plotted to what is called "X-Y-plot" in the QCA community with the simple plot() command for data exploration and hypothesis evaluation.

QCAtools does not provide the Boolean minimization inference algorithm, since this is already implemented in the QCA Package. It is designed as an addition to this package and fully compatible with its inputs and outputs.

Find it on CRAN.

python: mathdown

Mathdown merges markdown and Latex's math mode syntax in order to provide a simple and fast way for mathematical notetaking. You can write a markdown file as usual and simply use Latex's math commands without any further declaration (like $ signs) in the text. Mathdown has heuristics to detect this, interprets it as math and outputs a nice .pdf as well as a .tex files from your notes.

The source code and installation and usage instructions can be found under http://gitlab.points-of-interest.cc/jlewando/mathdown. I plan to convert the abandoned and somewhat buggy BASH/python version to R at some point.

R package: QCAsensitivity (beta)

QCAsensitivity is an R package to evaluate to robustness/sensitivity of obtained solutions/hypotheses in Qualitative Comparative Analysis (QCA), a qualitative, but formal research approach based on Boolean algebra.

The package is more of a playground to pin down and evaluate theoretical approaches I've been coming up with to analyse sensitivity/robustness of QCAs. Currently, the software is incomplete and hardly documented, but already useful to some extent. You can find it at http://gitlab.points-of-interest.cc/jlewando/qcasensitivity. The only person who has ever used this successfully is probably @beimwort.

MATLAB: arg -- Distributions of times to MRCAs in the Ancestral Recombination Graph

arg is a collection of Octave/MATLAB scripts to compute numerically as well as simulate distribution functions for times to most recent common ancestors in the specific population genetics model introduced as "Ancestral Recombination Graph" (ARG) by Griffiths 1991. It was developed in the course of my B.Sc. thesis "Distributions of times to most recent common ancestors in coalescents with recombination" in mathematics at Technische Universitaet Berlin (Berlin Institute of Technology).

The software together with installation and usage instructions can be found under http://gitlab.points-of-interest.cc/math-group/arg. For details on the ARG model and the concepts of times to most recent common ancestors, refer to the thesis text included in the repository's documentation folder.