Never trust a physicist on software. We have a special term at work, “Scientific Python”. When we say it to a physicist, they wear the hat proudly, but what we mean is “incomprehensible garbage that breaks as soon as you give it any input other than the one thing they tested it with”. If they want you to maintain it, the first step is to ask them to run it, look at whatever plot it generates, and write something from scratch that produces the same plot.
Never trust a physicist
Never trust a physicist on software. We have a special term at work, “Scientific Python”. When we say it to a physicist, they wear the hat proudly, but what we mean is “incomprehensible garbage that breaks as soon as you give it any input other than the one thing they tested it with”. If they want you to maintain it, the first step is to ask them to run it, look at whatever plot it generates, and write something from scratch that produces the same plot.
Yeah, they don’t even trust themselves!
Also just looking for falsification…