How do I start learning data science with Python?
Step: 1 - Setting up your machine
It's time to set up your machine now that you've made your decision. The simplest method is to simply download Anaconda from Continuum.io. It comes with the majority of the items you will ever require. The main disadvantage of using this approach is that you will have to wait for Continuum to update their packages, even if an update to the underlying libraries is available. If you're a beginner, it shouldn't matter.
If you have any problems installing, you can find more thorough instructions for different operating systems here.

Step: 2 - Learn the basics of the Python language
Begin by learning the fundamentals of the language, libraries, and data structure. One of the greatest places to begin your adventure is with Analytics Vidhya's free Python training. This course will teach you how to get started with Python for data science, and by the conclusion, you should be familiar with the language's fundamental concepts.

Step:3 - Learn Regular Expressions in Python
They'll come in handy for data purification, especially if you're working with text data. Going through the Google lesson and keeping this cheat sheet available is the best method to learn Regular expressions.

Step: 4 - Learn Scientific libraries in Python – NumPy, SciPy, Matplotlib, and Pandas
This is where the good times begin! Here's a quick rundown of some of the most popular libraries. Let's get started with some basic operations. Complete the NumPy tutorial, especially the NumPy arrays section. This will serve as a solid foundation for the future.
After that, take a look now at SciPy tutorials. Go over the introductions and the essentials first, then move on to the following sections based on your needs.
If you guessed Graph tutorials as the next step, you're mistaken! They're far too extensive for what we're looking for. Instead, scroll down to Line 68 in this python notebook (i.e. till animations)
Let us now turn our attention to pandas. Pandas is a Python package that provides Data Frame capabilities (similar to R). You should also spend a lot of time practicing here. Pandas have become the most powerful tool for all types of mid-sized data analysis.