Python: How to Begin Programming?

[Indy Coug]

So I'm watching some YouTube videos on Python and I'm getting increasingly interested in using this as a tool. For the CS members that are already experienced users, what are your suggestions for getting Python up and running?

Code base?
Libraries?
Developer tools?

[pelagius]

So I'm watching some YouTube videos on Python and I'm getting increasingly interested in using this as a tool. For the CS members that are already experienced users, what are your suggestions for getting Python up and running?

Code base?
Libraries?
Developer tools?

For general programming or data analysis programming?

[Indy Coug]

Both.

[Jeff Lebowski]

sudo apt-get install python2.7

[pelagius]

Both.

Anaconda Python Distribution

If you want to do any data analysis type programming (and it sets you up just fine for general programming) use Anaconda:

• Windows platform: anaconda (it's free ... https://store.continuum.io/cshop/anaconda/ )
  
  
• Unix/Linux/BSD (including Mac OX): probably anaconda (although you do have more options)

Anaconda automatically installs not only base python but also the defacto standard scientific and data libraries. For you the most important library is,

• Pandas: http://pandas.pydata.org/ (the documentation and tutorials are pretty good. Python for Data Analysis is the pseudo-canoncial book but it's little old given the pace of pandas development).
  
  
• Pandas is primarily built on the following lower level libraries: numpy (scientific array library ... like Matlab), scipy, and matplotlib.

General programming:

• Pythons online docs and tutorials are pretty good: https://docs.python.org/3/
  
  
• Python also has a big video library from the PyCons. Check out some of the Python 101 type videos: http://pyvideo.org/category
  
  
• Introductory Books: I like Quick Pythoyn and Python: The Essential Reference
  
  
• Stackoverflow is very active for python

Data Programming:

• Pandas Docs: http://pandas.pydata.org/pandas-docs/version/0.16.2/
  
  
• Video: PyData Conferences: https://www.youtube.com/user/PyDataTV/videos
  
  
• Books: Python for Data Analysis, Python for Finance, and High Performance Python
  
  
• Stackoverflow: http://stackoverflow.com/questions/tagged/pandas

Note

You can do either python 2.7 or 3.4 in the scientific/data python world. The scientific/data world has been pretty good about porting to python 3

[Indy Coug]

Excellent information. Thanks for taking the time to reply.

[pelagius]

Excellent information. Thanks for taking the time to reply.

Also, one of the things I did improve my python/pandas skills after initially learning it was to answer questions on stackoverflow. I answered questions for like 1.5 months

[YOhio]

I only use PyDev as my IDE for Eclipse. That's how I roll.

[Uncle Ted]

I only use PyDev as my IDE for Eclipse. That's how I roll.

It's python on the command-line with vim for the editor for me. I don't need no stink'n IDE.

Homebrew is your friend when it comes to installing modules, etc. on OS X.

[pelagius]

I only use PyDev as my IDE for Eclipse. That's how I roll.

That reminds me Indy.

I think most people in the data analysis/data scientist space are basically using the iPython Notebook as their IDE: http://ipython.org/ (it will be automatically installed with anaconda). Although, I think on windows some are still using the python plugin for Microsoft Visual Studio.

[BigFatMeanie]

I don't often use python, but when I do, I prefer vim. Stay vimmy, my friends.

[old_gregg]

lol @ you dorks

[Indy Coug]

It's python on the command-line with vim for the editor for me. I don't need no stink'n IDE.

Homebrew is your friend when it comes to installing modules, etc. on OS X.

I spent (wasted) 5 years of my life doing Pascal software development using vim in a UNIX environment. I guess at least I didn't have to worry about punch cards.

[Indy Coug]

Does anyone use Pycharm as their IDE? Thoughts? Which other IDEs are worth considering?