At one of my previous companies we hired a really intelligent (generally) and very likeable data scientist who arrived and wrote the worst Python I've seen. Actually it wasn't the worst, because it was all inline copy/paste code rather than convoluted OOPish code.
He could solve very complex problems, but his tooling was horrible. It wasn't his fault. He had solid education from a German STEM uni (with PhD.), but there was a serious lack of programming skill.
It would seem that because Python is "so easy" to get started that people don't feel they have to bother with learning any real programming skills beyond solving their immediate problem.
I don't blame this on the scientists; software is not their domain. The problem is with PHBs who don't know better and who make decisions based on the toolset used by the "special" people.
He could solve very complex problems, but his tooling was horrible. It wasn't his fault. He had solid education from a German STEM uni (with PhD.), but there was a serious lack of programming skill.
It would seem that because Python is "so easy" to get started that people don't feel they have to bother with learning any real programming skills beyond solving their immediate problem.
I don't blame this on the scientists; software is not their domain. The problem is with PHBs who don't know better and who make decisions based on the toolset used by the "special" people.