Publications by Econometrics and Free Software
Using explainability methods to understand (some part) of the spread of COVID-19 in a landlocked country
This blog post is based on an article I’m currently working on which you can find here. Contributions more than welcome! “You expect me to read all this?” The gist of this blog post can be summarised in the following sentence: lagged positive cases of the neighbouring regions of Luxembourg predict weekly positive cases in Luxembourg. But p...
9552 sym R (1734 sym/2 pcs) 14 img
How to treat as many files as fit on your hard disk without loops (sorta) nor running out of memory all the while being as lazy as possible
tl;dr This blog post is going to be long, and deal with many topics. But I think you’re going to enjoy it. So get a hot beverage and relax. Take the time to read. We don’t take enough time to read anymore. It’s a shame. But if you’re really busy, the tl;dr is that I found out a way of combining tidy evaluation and functional programming t...
10198 sym R (11160 sym/28 pcs) 12 img
The link between keyboard layouts and typing speed – Data collection phase
I’m curious about different keyboard layouts, and how that correlates with typing speed (if at all). I prepared a little 2 minute survey, and would be very grateful if you could take it. You can find it on this link. There’s no R code in this blog post, but I’ll be analyzing the data using R, promise 🙂 Hope you enjoyed! If you found this...
996 sym 4 img
Dealing with non-representative samples with post-stratification
Let’s go back to stats 101: what do you do if you want to know how many people like to play bingo in a certain population? The answer, of course, is to ask a sample of people if they enjoy playing bingo, compute the proportion and then… we’re done! Right? Well not exactly. This works if your sample is representative, which in practice, is n...
7059 sym R (4848 sym/16 pcs) 6 img
Building your own knitr compile farm on your Raspberry Pi with {plumber}
Rage is my fuel I’ve had the {plumber} package on my radar for quite some time, but never tried it. However, a couple of weeks ago, I finally had a reason to try it out and see how the package works. One of my main problems in life is that my work laptop runs Windows, and my second problem is that I need to compile \(\LaTeX\) documents (via Rma...
7850 sym R (2272 sym/9 pcs) 6 img
Is it worth the weight?
Intro Oh man, I did it again. Grab a coffee, this is going to be a long one. Weights got me confused. The justification for using weights seems simple enough; if you’re working with a sample in which one (or more) strata are over(under)-represented, you should compute weighted univariate statistics. I’ve discussed this already here. But what ...
5632 sym R (11043 sym/20 pcs) 10 img
The quest for fast(er?) row-oriented workflows
The past few weeks I have been exploring the speed of R. It started with this video in which I explained that R is not necessarily slower than any other interpreted language, as long as you’re using the built-in, optimized functions. However should you write your own implementation of an algorithm, especially if that algorithm requires the use ...
9449 sym R (10572 sym/21 pcs) 10 img
Speedrunning row-oriented workflows
If you haven’t, you should read this first. This is part two. Speedrunning is the… hrm… – sport? art? – of playing games from start to finish as fast as possible. Speedrunning requires an insane amount of knowledge of the game being played, as well as an enourmous amount of skill. Also, contrary to what you might think, it is a communit...
5344 sym R (3222 sym/8 pcs) 6 img
How to write code that returns (Rmarkdown) code
One of the most useful aspects of using a programming language instead of… well, not using a programming language, is that you can write code in a way that minimizes, and ideally, eliminates the need to repeat yourself. For instance, you can write a function to show you a frequency table, like so: suppressMessages(library(dplyr)) create_table ...
5406 sym R (3083 sym/14 pcs) 4 img 1 tbl
Add logging to your functions using my newest package `{loud}`
This is a short blog post to announce the early alpha, hyper unstable, use at your own peril, package I’ve been working on for the past 6 hours or so (actually longer if I add all the research/study time). This package provides the function loudly() which allows you to do cool stuff like: # First two lines install the package # install.packages...
3126 sym R (4404 sym/9 pcs) 4 img