I am committed to improving reproducibility in scientific research. Because of that, I am actively involved in the development of software to facilitate this endeavor. Here are some R packages I have co-developed with the Evolutionary Ecology Group at the University of Cambridge.
Easy reproducible research on git
The aim of resrepo is to encourage and facilitate good practices when setting up and managing git repositories for scientific research projects. It also provides a template for a tidy repository structure that can be used for any project, with functions that help keep a clear separation of code, data and results. It also provides functionality to help manage large data that cannot be tracked on a git repository.
Walking through the geographic space using graphs
geoGraph aims at implementing graph approaches for geographic data. In geoGraph, a given geographic area is modeled by a regular grid, where each vertex has a set of spatial coordinates and a set of attributes. 'Traveling' within the geographic area can then be easily modeled as moving between connected vertices. The cost of moving from one vertex to another can be defined according to attribute values, which allows for instance to define friction routes based on climate.
Clustering for population genetics in R
tidygenclust provides functions and methods to run genetic clustering in R, using the commonly used ADMIXTURE program as well as the python package fastmixture. It also helps to align and compare multiple runs of the same or different K using the functionalities of the python package clumppling.
Tools for working with map projections in R
The goal of crstools is to facilitate working with map projections (technically coordinate reference systems, CRS) in R. We provide functions to choose the appropriate map projection for a given application, visualise the resulting distortion, and georeference data from unknown projections. The package seamlessly interfaces with the popular sf and terra packages for spatial data handling.