Your Complete Python Roadmap to Become a Data Scientist

0 Warm‑up

Before you start, watch the 30‑minute PyCon talk by DataRobot founder Jeremy to understand why Python is useful.

1 Set Up Your Computer

Download Anaconda from Continuum.io, which bundles most of the tools you’ll need for Python data science.

Installation instructions for different operating systems are available at the DataRobot blog.

https://store.continuum.io/cshop/anaconda/

http://www.datarobot.com/blog/getting-up-and-running-with-python/

2 Learn Basic Python Knowledge

Start with the Codecademy Python track to learn language basics, libraries, and data structures.

http://www.codecademy.com/tracks/python

After completing the tutorial you should be able to write small programs and understand classes and objects. Focus on Lists, Tuples, Dictionaries.

Complete the exercises on HackerRank to reinforce learning.

Alternative: Google’s two‑day Python course.

https://developers.google.com/edu/python/

3 Learn Regular Expressions

Regular expressions are essential for data cleaning, especially text data. Follow the Google regular‑expression course.

https://developers.google.com/edu/python/regular-expressions

Keep a cheat‑sheet handy, e.g., www.debuggex.com/cheatsheet/regex/python.

Complete the “baby names” exercise.

https://developers.google.com/edu/python/exercises/baby-names

For more practice, see the text‑data‑cleaning course.

http://www.analyticsvidhya.com/blog/2014/11/text-data-cleaning-steps-python/

4 Learn Python Scientific Libraries

Explore NumPy, SciPy, Matplotlib, and Pandas.

NumPy – practice array operations.

http://wiki.scipy.org/Tentative_NumPy_Tutorial

SciPy – start with the introductory tutorial.

http://docs.scipy.org/doc/scipy/reference/tutorial/

Matplotlib – the IPython notebook up to line 68 (animations) is sufficient.

http://nbviewer.ipython.org/github/jrjohansson/scientific-python-lectures/blob/master/Lecture-4-Matplotlib.ipynb

Pandas – learn DataFrames, start with the 10‑minute intro then deeper tutorials.

http://pandas.pydata.org/pandas-docs/stable/10min.html

http://www.gregreda.com/2013/10/26/intro-to-pandas-data-structures/

Additional Pandas resources: Python for Data Analysis by Wes McKinney, and the official tutorials.

http://pandas.pydata.org/pandas-docs/stable/tutorials.html

5 Effective Data Visualization

Complete the Harvard CS109 visualization assignment.

http://cm.dce.harvard.edu/2015/01/14328/L03/screen_H264LargeTalkingHead-16x9.shtml

6 Learn Scikit‑learn and Machine Learning

Study Harvard CS109 lectures 10‑18 to cover supervised (regression, decision trees, ensemble methods) and unsupervised (clustering) algorithms.

http://cs109.github.io/2014/pages/schedule.html

Recommended reading: "Programming Collective Intelligence".

Also consider Andrew Ng’s Machine Learning course on Coursera.

https://www.coursera.org/course/ml

Practice with the Kaggle data‑science challenge.

http://www.kaggle.com/c/data-science-london-scikit-learn

7 Practice, Practice, Practice

Compete in live Kaggle competitions to apply everything you have learned.

http://www.kaggle.com/

8 Deep Learning

After mastering most machine‑learning techniques, explore deep learning. A good introductory article is linked below.

http://www.analyticsvidhya.com/blog/2014/06/deep-learning-attention/

For comprehensive resources, visit deeplearning.net for lectures, datasets, challenges, and tutorials.

http://deeplearning.net

Learn the basics of neural networks from Geoffrey Hinton’s Coursera course.

https://www.coursera.org/course/neuralnets