colinpmillar
commented
4 years ago The ... [assessment] code is in github following the TAF structure.
However, I’ve to say that the assessment process is usually not as clean and simple as it seems or as it is in some cases. In the data compilation process there are always problems that make you writing quick and dirty code, some countries not uploading their data into intercatch, other making historical data revisions…. In the model fit, you usually have to test different model settings, they take long time to run, you have to use a cluster or similar because otherwise your PC crashes…. And the analysis, in my case, the code is so long and not very friendly.
So finally, you end up with many R scripts, many intermediate ouputs….
But everything is there and is transparent and reproducible.
Next year I’ll continue working in github following the TAF guidelines and It’ll get better.
summary
Just an idea, as we do get potentially uefullfeedback via email, and it would be good to collate somewhere.
related issues