Having Streamlit Superpowers: The Best GPT-4 Prompts For Guaranteed Data Visuals

Author(s): John Loewen, PhD

Originally published on Towards AI.

No-fuss prompting for error-free Python Streamlit data viz code

How easy is it to create data visuals with GPT-4 and the Python Streamlit library?

Way easier than I ever expected!

A little bit of fancy footwork with a good dataset and we can rattle off chart after chart (even maps) giving us in-depth analysis and stories for our data set.

What kind of charts and maps you ask? We can prompt GPT-4 for:

Time-series analysisComparative analysisDistribution analysisInteractive global choropleth map

Let me take you through the simple steps to accomplish this — with a minimal amount of fuss.

Personally, I like to do all of my chart generation using publicly accessible global development data.

For this example, let’s use the UNESCO global democracy index dataset that can be downloaded HERE.

Once we have a CSV file from the UNESCO site in our hands, we can upload this file to GPT-4 and prompt away:

Prompting GPT-4o to give us a description of the field in the Global Democracy Index dataset

Response from GPT-4:

This dataset contains 6 fields across 32,518 rows. Here’s a concise overview of each field:

Entity — Name of the country or region.Code — ISO 3-letter country code (some entries are missing).Year — The year of the observation.Electoral democracy index — A numerical score representing the level of electoral… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI