The best place to install Python and its many useful libraries is Anaconda.

If you are more comfortable with a Mathematica-like setup, use IPython to run your code. If you like Matlab then Spyder will be the closest development environment. Both of them come with Anaconda.

If you want to learn how to use Python for Econometrics, Statistics and Data Analysis, here is a good introductory book. There are many more available on Amazon.

If you are interested in using Python for Machine Learning/AI projects, here are some wonderful resources

Scikit-learn is a collection of very useful Python tools for data analysis

PyMC is designed for Bayesian statistical models.

Shogun can be used if you want to work on Support Vector Machines