If you ever want to get serious about data science soon or later you’re going to have your hands on some Python code.
If you are like me – coming from the JVM world – you probably think “yeah, Python … should be cool!!”. Everybody is using it, the syntax looks concise, and the machine learning ecosystem is pretty dense in python: theano, neon, scikit-learn, …
..So yeah let’s get started… and if you’ve never written any Python code before I’ll tell you it’s not going to be that fun.
For sure I know it’s different language. It’s a dynamic language, you’ve got to get the indentation right and the way to call the main method is even worse than Java
if __name__ == '__main__': main()
but that’s not the problem, you can get all of these quite right. The problem starts when you dig head first into numpy. For sure it’s a powerful library and the syntax is very concise … but you have to learn it … there are cases where there’s no way to guess what it does
a = a[:,:,::-1]
See, and people complain that Scala has too much ascii-art!!!
Well this line looks cryptic, for sure, but once you’ve learned this tricky syntax you realise how powerful it is and how many lines of code it saves.
So if you’re like me, watch a couple a python videos before you get started (or even better sign up for an online course) and I’m sure you’ll enjoy it much more than I did the first time.