Experiential Workshop on Python for Data Science
May 10-12, 2023
virtually cohosted by Namal University Mianwali, BZU Multan, and IBA Karachi
Venue: Namal University, Mianwali
Session Plan
Day 1
Python fundamentals and data gathering
Morning
An introduction to the Jupyter IDE and fundamental types in Python. Students will learn how to:
Write clear, well-documented Python code in Google Colab.
Perform calculations using the math library.
Extract information from deeply nested lists and dictionaries.
Afternoon
How to obtain data from the web. Students will learn:
What an API is, and the increasingly broad range of APIs available.
How to start using an unfamiliar API by reading and understanding the documentation.
How to use the requests and pandas libraries to obtain and explore data from APIs.
Exercises will include: obtaining live share prices from the Alphavantage API, and covid statistics from the Johns Hopkins API.
Day 2
Data exploration and visualisation
Morning
How to explore and analyse large, complex datasets with the pandas library. Students will learn:
To work with and manipulate DataFrames.
How to clean and check the quality of datasets.
To compute descriptive statistics, including correlation coefficients.
Exercises will include: Exploring gender pay gap data and crime data.
Afternoon
How to produce charts and visualizations to aid data exploration. Students will learn:
How to produce and interpret static charts using matplotlib and seaborn.
How to make interactive charts and dashboards using plotly.
Exercises will include: Determining the strongest predictors of house prices.
Day 3
Working with unstructured data
Morning
The basics of natural language processing. Students will learn:
How to clean unstructured text data using the nltk library and techniques including tokenization, stopword removal, stemming and lemmatization
Exercises will include analysing British political speeches
Afternoon
Questions and troubleshooting. A chance for students to ask for advice on project work, get help with debugging their code, or ask any other burning Python questions.