>>13012543>Why all these new age grad and even post doc ride the python cock so hard?>Not everyone doing research has to be a coding sweat.holy shit do you think python is coding sweat vs matlab?
I'm a cell biologist turned computational chemist (mostly generative models for drug design + knowledge networks).
I used R + matlab in grad school, python 95% since.
1) Matlab is absolutely amazing for it's image processing. I loved that shit. Python is....somewhat blocky in comparison.
2) Holy shit I hate the 135146 different and unencessary data types in Matlab.
3) Instead of a "base" language, it's all libraries integrated into libraries. Matlab excels at adding base library functions, BUT. Every goddamn 2 years my code breaks because they absorbed some function into another one. This is the real killer of matlab. They develop amazing functions/library types and then kill them off in 4 years by absorbing/succession with other shit. And they release every 2 years. I either code in the strengths (library whatever functions, I forget what matlab calls their packages) and expect it to break ever few years, or code in their base-base-base language and hate everything about it.
4) I dunno what you're smoking by python is easy as fuck. R is best for statistics, but python is the goddamn ML king. Pytorch > tensorfow, btw. Tensorflow was good, pytorch became better, tensorflow 2 with integrated keras tried to copy pytorch, but pytorch is just straight pythonic (like fucking how, it's a great feat) where tensorflow is not. Babby's first ML models are usually in tensorflow since it's so goddamn easy to implement, but serious ML models generally do much better with pytorch.
Python is just a great first language. /g/ is full of non-important high-schoolers and undergrads who think the closer to assembly the better.
