license: cc0-1.0
pretty_name: Open Data Notes (synthetic city-service tables)
language:
- en
tags:
- synthetic
- tabular
- csv
- teaching
size_categories:
- 1K<n<10K
configs:
- config_name: bus_stop_accessibility_2025
data_files: data/bus_stop_accessibility_2025.csv
- config_name: library_opening_hours_2026
data_files: data/library_opening_hours_2026.csv
- config_name: street_tree_inventory_sample_2024
data_files: data/street_tree_inventory_sample_2024.csv
Open Data Notes: synthetic city-service tables
Every value in this dataset is synthetic. The tables describe a fictional city. No real city, place, survey, organisation or person is described, and the numbers must not be used as facts about anywhere. They are made for teaching: data cleaning, joins, grouping and simple charts.
| table | rows | columns |
|---|---|---|
bus_stop_accessibility_2025.csv |
1,240 | stop id, name, district, x/y in metres, shelter, bench, step-free access, tactile paving, audio announcements, survey date |
library_opening_hours_2026.csv |
38 | branch id, name, district, opening hours for each day of the week |
street_tree_inventory_sample_2024.csv |
5,000 | tree id, species (common and Latin), district, street, x/y in metres, year planted, trunk diameter, height, condition |
Coordinates are metres on a local grid, not latitude and longitude. Species names are real tree species; everything about the trees is generated.
How it was made
generate.py builds all three files from a fixed seed with the Python standard library only.
Running python generate.py reproduces data/ byte for byte.
Companion site
The same tables are described at https://www.arthsingh.com/open-data. That site is part of a research project on automated web agents: it records visits (time, path, request headers, a keyed hash of the IP address), as its privacy page explains. It does not set cookies or store full IP addresses.
Licence
CC0 1.0. No attribution required.