I am curious about most things Stats/Data/Coding and currently working as research associate at the MRC Biostatistics Unit, University of Cambridge, UK. I am doing research on the optimal design of clinical trials, and genetic patterns liked to the recovery after Traumatic Brain Injury.
Got an evening to spare? Why not philosophize about the differences between Frequentist and Bayesian Statistics? My go-to conversation starter would be The Lady Tasting Tea by David Salsburg.
If R is the ‘English’ of coding, Julia would be its ‘French’! Real-world statistical analyses quickly require hundreds of lines of code and having the right tool for the job is paramount. The impact of Hadley Wickham, the tidyverse, and RStudio for making data science sexy and accessible can hardly be overstated. Julia, however, is much more elegant when heavy lifting becomes involved: true source-to-source automatic differentiation, syntactic loop fusion, and C-like speed.
I am passionate about reproducible and open research. I consider git a key element in achieving reproducibility. It has become so much more than a version control system - especially when adding the ‘wider ecosystem’ of GitHub/Lab and the like to the equation. Too bad, it is a nightmare to learn, the true Latin of data science so to speak… Please, also check out the Turing Way, a great initiative of the Alan Turing Institute to raise awareness and to promote best practices around reproducible research.