Using Generative AI to Create Datasets for Graph Algorithms
The best way to quickly understand graph algorithms is to have high-quality datasets to help you visualize how the graph algorithms work step-by-step.
This section demonstrates how to use generative AI to generate data sets.
Sample Prompt
Generate a table of the 20 largest metro regions in the US.
Each row has a metro region.
Each row has the following columns:
1. Rank
2. Name
3. Population
5. Latitude
4. Longitude
Round all numbers to three significant places.
If you are happy with the results you can issue the following prompt:
Return a downloadable JSON file of this table.
US Cities Graph Network Demo
The following demo walks you through the process of creating a sample data set of the top 20 metro regions in the US. This dataset is ideal for testing various algorithms such as search, path and traveling salesperson.