part of one of my previous jobs involved converting code from MatLab to Scala. Needless to say, it wasn't very fun. Languages like Matlab, Stata, SAS, or R don

part of one of my previous jobs involved converting code from MatLab to Scala. Needless to say, it wasn't very fun. Languages like Matlab, Stata, SAS, or R don't seem to be designed for professional developers. They tend to break all the assumptions that every other languages have (like array indices beginning with 0). I also feel this way about Pandas, it seems like it was designed by a data scientist and not a developer. Because the people using these particular languages don't care about elegance (they usually just want the graphs or papers or whatever) and that no one is writing commercial systems that need to be maintained, there's no real incentive for them to design a language that makes any sense. Making things ad-hoc and less axiomatic does have its benefits however, these kinds of languages can be surprisingly expressive and powerful.