Mapping PyData

I created this project out of a desire for a better map of the international PyData community than the one available on Meetup. It's since grown to cover Python user groups more broadly through the PSF's Meetup Pro Network too (see "Networks" below).

This repo consists of three main parts:

• PyDataMap.py a script which scrapes Meetup.com for group and event data for each network in NETWORKS (currently PyData and the PSF Python Network), populates the geocode_cache.json with each group's location (preferring the city Meetup reports for the group, falling back to parsing the group's name when the city is missing or a placeholder), updates each network's own CSV (pydata_groups.csv, psf_groups.csv) which stores data about upcoming and past events in aggregate by group name, and then produces static versions of the maps for each network plus a combined map across all of them. This script is scheduled to run daily. Groups that exist outside of Meetup (university clubs, conference series, Discord-only communities, groups that have moved to LinkedIn or their own website) can be added manually to a network's manual CSV (e.g. pydata_groups_manual.csv) and are merged into the maps at render time, or flagged directly on their existing row via the non_meetup column (see "Non-Meetup groups" below).

• MapsExplained.py a marimo notebook which can be used to create and explore the PyData maps based on cached data in geocode_cache.json and pydata_groups.csv. This is intended to make it easy to create your own maps with this data. It also includes some example queries that can be made against the collected data i.e. top 10 most recent events.

• The maps (which are hosted via GitHub Pages), see the full index below.

Map index

All maps are static HTML, regenerated daily by PyDataMap.py, and hosted via GitHub Pages.

Network	World map	Active	Inactive	Non-Pro
PyData	pydata_world_map.html	pydata_world_map_active.html	pydata_world_map_inactive.html	pydata_world_map_non_pro.html
PSF Python Network	psf_world_map.html	psf_world_map_active.html	psf_world_map_inactive.html	psf_world_map_non_pro.html
Combined (all networks)	all_python_world_map.html	all_python_world_map_active.html	all_python_world_map_inactive.html	n/a

• World map is a 1:1 recreation of the relevant Meetup Pro network page (PyData, PSF) but with locations corrected.
• Active is intended to make it easy to spot active groups with a view towards attending / speaking at them.
• Inactive draws attention to groups that haven't hosted an event in a while. Consider volunteering for, speaking at, or sponsoring these groups to help them out.
• Non-Pro shows only groups outside that network's Meetup Pro network (manually-added groups, see "Non-Meetup groups" below). There's no combined Non-Pro map: "in the Pro network" is a per-network fact, so it doesn't have one unambiguous meaning across networks.
• Combined merges every group from every network onto one map. A group listed in more than one network still only gets a single marker (see "Networks" below).

Also in this repo but not part of the regularly-regenerated set above: experiments/TourDePyData contains a personal, one-off map tracking which UK/Ireland PyData groups Hugh has spoken at as part of his Tour de PyData challenge.

Networks

Meetup Pro networks (PyData, the PSF Python Network, and any future ones) are declared in the NETWORKS dict at the top of PyDataMap.py: each entry just needs a Meetup Pro URL, a CSV filename, an optional manual CSV, a sanity-check minimum group count, and a filename prefix for its maps. Adding a new network means adding an entry there; the scraping, geocoding, enrichment, and map-generation code is shared and network-agnostic.

Each network gets its own CSV and its own set of maps, so nothing about the existing PyData maps/URLs changes. A network column on each row records which network's CSV a group came from (defaults to that network if the column is missing, for backwards compatibility with CSVs written before multi-network support).

For the combined maps, groups are merged by Meetup URL: a group present in more than one network's CSV becomes a single row with a networks field listing all of them (e.g. pydata,psf) rather than being duplicated into two markers at the same spot or silently dropped from one. Where a group is genuinely cross-listed, its popup on the combined map shows a small badge listing which networks it belongs to; single-network groups don't show this, since the map's own title already makes that obvious.

FAQ

Why is my group shown in the wrong location?

I am using geopy to geocode each group's location, preferring the city Meetup reports for the group and falling back to parsing a location out of the group name when the city is missing or a placeholder like "New group". This is not a perfect process, particularly for groups whose city doesn't resolve cleanly on its own (ambiguous names shared with other places, unusual formatting, etc.) or whose name-based fallback doesn't follow the PyData {location} convention. To get around this I've added a series of aliases (or hints) to the geocode cache as shown below.

"hints": {
    "PyMC Online Meetup": null,
    "PyData En Espa\u00f1ol Global.": null,
    "NEO AI - a PyData Group": "Cleveland, Ohio, USA",
    "PyData Ireland": "Dublin, Ireland",
    "PyData T&T": "Port of Spain, Trinidad and Tobago",
    "PyData Katsina": "Katsina, Nigeria",
    "Copenhagen Julia Meetup Group": "Copenhagen, Denmark",
    "PyData Boston - Cambridge": "Boston, Massachusetts, USA",
    "PyData Athens": "Athens, Greece",
    "Pydata Belgium": "Brussels, Belgium",
    ...

If your group is shown in the wrong location you can raise an issue or a PR with an alias for your group (or manually edit the latitude and longitude in the coords section of the cache file).

How can I make versions of this map focused on a specific country or area?

The maps support URL hash navigation in the format #zoom/lat/lng. Simply append a hash to any map URL to set the initial view.

Examples:

• The UK: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#6/54.5/-2
• Europe: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#4/50/10
• Continental US: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#4/39/-98
• India: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#5/20/78
• Australia: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#4/-25/135
• Brazil: https://hevansdev.github.io/mapping-pydata/pydata_world_map_active.html#4/-15/-50

To find coordinates for your own view:

1. Open any of the maps
2. Pan and zoom to your desired view
3. The URL hash will automatically update as you move
4. Copy the URL to share

Contributing

If you have an idea for how to improve this project please fork and raise PRs. Contact Hugh for all other inquiries.

Adding a group that isn't on Meetup

If you know of a PyData community that doesn't have a Meetup page (a university club, a conference series, a Discord-only group, etc.), you can add it to pydata_groups_manual.csv (the manual CSV for the pydata network, set via manual_csv_file in NETWORKS). The required columns are name, url, city, country, lat, and lon. Use the source column to describe where the group comes from (current values are discord, conference, and university). All other columns are optional and can be left blank. The PSF network doesn't have a manual CSV yet, so add a manual_csv_file entry for psf in NETWORKS and create the file in the same shape if that's needed.

Non-Meetup groups

Meetup's "has upcoming events" / "days since last event" fields drive the Active and Inactive maps, but that data doesn't exist for groups that never had a Meetup page, or that have since moved off Meetup (e.g. to LinkedIn or their own website). Rather than showing those groups as misleadingly active or inactive, set the non_meetup column to True on their row and they'll render with the same neutral marker used on the simple World Map instead.

• Groups added via pydata_groups_manual.csv default to non_meetup = True automatically, since they have no Meetup activity data by definition.
• Groups that already have a row in pydata_groups.csv (because they used to be tracked on Meetup) but have since moved elsewhere: update their url to the new LinkedIn/website page and set non_meetup to True on that same row. There's no need to move them to the manual CSV or duplicate the entry.

Any row that leaves non_meetup blank is treated as False (normal Meetup-tracked group).

Cataloguing multiple links for the same group

A group only ever gets one row and one marker, but it can have more than one link worth keeping: a LinkedIn page it's also active on, a Discord, an old Meetup page it's moved off of, etc. Use the alt_urls column on its existing row in pydata_groups.csv (or pydata_groups_manual.csv) for this rather than adding a second row, which would just create a duplicate marker at (roughly) the same spot.

To add an alt link: find the group's row and set alt_urls to the link, optionally labelled as Label::https://example.com. An unlabelled bare URL falls back to a label derived from its domain. To add more than one, separate them with |.

alt_urls: LinkedIn::https://www.linkedin.com/company/pydata-cyprus

alt_urls: LinkedIn::https://www.linkedin.com/company/pydata-cyprus|Discord::https://discord.gg/example

This is independent of non_meetup: a group can still be an active, Meetup-tracked group (non_meetup left blank/False) and simply have an extra link noted alongside it. Adding an alt_urls value doesn't change how it's styled on the Active/Inactive maps.

If a group has moved off Meetup entirely rather than just gaining a second presence, combine this with the non-Meetup workflow above: swap url to the new primary link, set non_meetup to True, and put the old Meetup link in alt_urls so it isn't lost:

url: https://www.linkedin.com/company/pydata-cyprus
non_meetup: True
alt_urls: Meetup (until 2026)::https://www.meetup.com/PyData-Cyprus

Either way, alt_urls renders as extra links in the marker popup on every map, underneath the group's primary link.