/2024_maps_supplement

Supplementing Datameet maps with some newer, updated ones for the 2024 elections

The UnlicenseUnlicense

2024_maps_supplement

Supplementing Datameet maps with some newer, updated ones for the 2024 elections

The maps in the Datameet repo meet most of my needs, but some areas like Assam and J&K still have pre-delimitation parliamentary seat borders.

This is an attempt to provide maps that can substitute some of the outdated ones in the Datameet repo.

These geojsons may not be research quality, but I think they're good enough for the purpose of visualising on websites.

Done

Notes

  • If you want to use these geojsons for visualisation on the web, its best to convert them to topojsons, they will be smaller in file size.
  • Source maps used for sikkim, assam, ladakh, j and k are in the source_maps folder
  • The main source maps were taken from this press note pdf from the election commission
    • The various maps in the pdf can be extracted as svgs if you want updated maps for other states/UTs.
  • Other source maps were taken from the websites of the state election commissions
  • Areas in J&K claimed by India but administered/controlled/occupied by others have been assigned the ls_seat_code value of '999'
    • Have created two India maps. The first one India Lok Sabha seat map 1 lets you decide how to deal with these other areas.
    • If you don't want to mess with anything and just want an India map with 543 features, one feature for each seat, use India Lok Sabha seat map 2
  • Because the maps were created by georeferencing images in QGIS, the international borders will be off, so use at your own risk.
    • The quality of the all-India Lok Sabha seat maps could be better, will see if I can improve them, no promises though.
    • Hopefully someone from the GIS community in India will step up and create a more accurate seat map that we can use instead. Will link to it here if anyone does.
  • If someone comes across this repo online, and wants to create a better map, see if you can build one from the taluk up for J&K and Assam.
    • Because even if you georeference the election commission images and create slightly better maps than the ones in this repo, it will still just represent marginal gains in accuracy.
    • Building lok sabha seats from the taluk up is time-consuming, but that should be the next step.
    • As a starting point, you can go through govt. gazette notifications (J&K ones are here). They have lists of which districts/taluks/tehsils of a state are in which lok sabha seat etc.

Corrections

  • Apr 20, 2024 -- Thanks to Srinivasan Ramani of Hindu for pointing this out, had interchanged the names and election commission ids of two lok sabha seats in Assam -- "Diphu (ex Autonomous District)" and "Kaziranga (ex Kaliabor)". They should be associated with the right seats now.

Update for J&K assembly map (Aug 9, 2024)

The map is based on this pdf from the J&K chief electoral officer's website and an official-ish geojson from this NIC map server. Hat-tip to Arun Ganesh aka planemad who pointed out the NIC map server in one of his gists. And if that govt pdf ever gets taken down for any reason, I've hosted it here too.

The geojson from the map server had actually mixed up the assembly seat numbers and names, but that's ok, i'm just thankful there's something out there with correct-ish borders. I've fixed the seat numbers and names using the government pdf.

(If you want the geojson to look like J&K as shown in the pdf, you'll need to do some extra work and 'reproject' it from EPSG:4326 to EPSG:7755.)

Now as the assembly election approaches, there's a chance that the election commission might change the seat numbers up, it's unlikely, but in case they do, will update it in this geojson as well.

I've tried to be a little more inclusive and generated Urdu and Hindi versions of all the seat names and district names using Claude Sonnet 3.5. This should help the national media and the local media in J&K a bit.

Most of the transliterations should be ok, but please cross-check and correct them before using them in production. (And if any of the spellings are drastically wrong, let me know by raising a github issue or mailing me on mail@shijith.com, and I'll fix them.)

And speaking of production, please don't use the geojson as is. Convert it into a topojson and use that, it'll be much smaller in filesize, your user will thank you :)

Lastly, there is feature numbered 9999 in the geojson, it's the part of J&K occupied/controlled/administered (take your pick) by Pakistan. If you want to show J&K as the Indian government defines it, you'll have to show that area too.