@sharoz I do have concerns about how easy it is to produce bad analyses with #python.
But I feel similarly about #rstats packages like #caret - you don't need to know what a classifier is, in order to train one.
This might be controversial, but we need to ask whether making it so easy is a good thing; and whether there is an argument to be made for building-in a small degree of difficulty.
Far from gatekeeping, it would be a measure against poor analyses that might cause actual harm.
New PolicyViz Podcast episode with yours truly!
New PolicyViz Podcast episode with yours truly!