SQL is not enough

As Data Scientists, we always have to roll up our sleeves to perform data processing/engineering jobs in data science pipelines.

It's widely accepted that SQL is core skill for performing ETL. However, I feel Python is equally important for its flexibility and thriving community, it's like Swiss Army Knife in your pocket!!

Every data scientist using Python would easily know how to apply Pandas, Numpy, Matplotlib, Seaborn, Plotly, Sklearn and other Stats modules on tabular data.

However, as Data Science field is converging with Software Engineering, I would add 10 more modules/tricks in the list to write production ready code.

1. Try/Except/Assertions in code to not break your pipeline
2. Collection module to extend the standard Python data types
3. OS for working with directory and Datatime for dates
4. Python Debugger module to avoid multiple print statements
5. Pylint for test cases and writing clean code
6. Any SQL ORM for handshake between Python and SQL databases
7. Decorators for wrapping pre and post checks around your functions
8. Modularize code with user defined functions and apply OOPs concepts
9. Lambda expressions and Map for quick data updates
10. GIT for code versioning and control

At times few things are very easy to do in SQL however I personally feel most of the times Python wins, it has extensive library of modules to perform complex data manipulations.

Share your favorite python modules to help others! Thank you!

#Python #Learning #DataScience #DataEngineering