Skip to content

My Jupyter notebooks in which I practice data science.

License

Notifications You must be signed in to change notification settings

arv-anshul/notebooks

Repository files navigation

Notebooks

Contains all my 📓 Notebooks where I have performed Data Analysis on unique datasets.

Important

See data/README.md to know about datasets I have used.

Directory Info

Perform analysis on YouTube Watch History data (exported via Google Takeout).

  • Previously using Pandas but switched to Polars as I started exploring it.

  • Used polars's amazing syntax to handle data, preprocess the text data and handle datetime data.
  • Plot many graphs to show some amazing insights present in data.
  • Build ML model to predict videos "Content Type" from its title.
  • Build a Channel Recommender System which recommends similar channels from channel's videos' title and tags.

Perform analysis on Spotify Streaming History data (exported via Spotify website).

  • Analysed data from the perspective of Track, Artist, Album, Playlist and Time.
  • Used polars builtin plot namespace (which uses hvplot library internally) to plot analysis graphs.

A project from CampusX's free course on Credit Risk Modeling by Rohan Azad.

Created a dashboard using Streamlit which fetches data from ECI official website.

  • Used httpx to fetch data asynchronously.
  • Used polars.LazyFrame to manipulate data efficiently.
  • Used streamlit to create dashboard.
Where is Notebooks?

There are no notebooks present in this project because I've converted those into .py scripts because I have to create a dashboard using it and converted notebook's non-async codes into async code.

Instead of Jupyter Notebook uses Marimo a new project as an alternative to Jupyter Notebooks. See marimo website.

🏟️ Try Marimo playground at marimo.app

This directory contains extra notebooks which are independent of each others. Created these notebooks just for learning or fun purpose.