DEDUPED Hot AI repos. Real use cases. What people build. Get the Monday 5

How to explore a dataframe with drag-and-drop charts in Jupyter with Pygwalker

Call pyg.walk on a dataframe to get a Tableau-like interface right in your notebook.

Short answer: Run pip install pygwalker, or conda install -c conda-forge pygwalker; then in a notebook, import pandas as pd and Pygwalker as pyg; load your data, for example df = pd.read_csv('./bike_sharing_dc.csv'). All steps.

Who it is forAn analyst who wants to look at a new dataset fast without writing matplotlib code for every chart.

The problemWriting plotting code for every question slows down early exploration of a dataset.

The outcomeAn interactive chart builder in your notebook, with filters, facets and a data profile table.

What you need first

  • Python with pandas, polars or pyarrow data
  • Jupyter Notebook, JupyterLab, Colab or Kaggle, or Streamlit for a web app
  • pip, conda or mamba to install

Steps

  1. Run pip install pygwalker, or conda install -c conda-forge pygwalker
  2. In a notebook, import pandas as pd and Pygwalker as pyg
  3. Load your data, for example df = pd.read_csv('./bike_sharing_dc.csv')
  4. Run walker = pyg.walk(df) and drag fields onto rows and columns to build charts
  5. Use the data table to see distributions, add filters or change data types

Prompt for your agent

Using Kanaries/pygwalker, open <csv file> in a Jupyter notebook with pyg.walk and set it up with spec_path so my charts are saved. Suggest three charts I should build first to understand this dataset.

Caveat. The UI runs inside the notebook, so it is for exploration rather than a finished report on its own.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for Kanaries/pygwalker and "Explore a dataframe with drag-and-drop charts in Jupyter"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer