Considering using C++ instead of python for data science, it is good idea?

For deployment, it's a great idea. For iteration/exploration/graphing/model prep/etc. which honestly take up the most time, horrible.

Pros: better language, significantly faster at runtime.

Cons: limited toolset, high cost of iteration, smaller community for support.

@Root @RememberMe I see.. thanks tho

@johnmelodyme if you really want to take up data science, the hard part is learning the maths and stuff behind it. Languages/deployment etc. is the easy part. Literally pick whatever/the simplest one and have at it.

Making a graph or visualization of a dataset is easy. Can you interpret what it means? From the graph and other metrics, can you figure out what you'd do to clean up the data or extract whatever insight you're looking for? That's hard.

@RememberMe Any links to learn this ? => the Mathematics and stufffff

@johnmelodyme while it's not ML, a great site that I used is https://d2l.ai. It goes through the math, and the implementations in python.