Bangladesh Election 2026 Interactive Maps

Note: If you have a better or more authoritative vote data source, you're welcome to contribute! Submit a pull request with updated data or scraping logic—I'll gladly review and merge improvements.

An interactive choropleth map showing constituency-level results from the 2026 Bangladesh general election across 300 parliamentary seats. Built with Leaflet.js on a dark CARTO basemap with ColorBrewer palettes.

Live site: smanahian.github.io/ElectionMap

Built by S M A Nahian (https://smanahian.com).

Features

Map layers

Layer	Description
BNP Vote Share	Bangladesh Nationalist Party (Blues)
Jamaat Vote Share	Bangladesh Jamaat-e-Islami (Greens)
11-Party Alliance	Eleven-party Islamist coalition led by Jamaat (Purples)
BNP vs 11-Party	Diverging red-yellow-blue map with sigmoid scaling
Winner Map	Color each seat by winning party
Turnout	Voter turnout percentage (Yellow-Green)
Victory Margin	Winner's margin of victory (Yellow-Red)

Interactive tools

• Single View / Compare View — top-level toggle between single-stat exploration and side-by-side comparison mode
• Constituency / District toggle — switch between individual seat view and aggregated district-level view with dissolved boundaries
• Scatter plot (Compare View) — plot any stat vs any stat with linear regression best-fit line, Pearson r and r² values. Click dots to select seats. Dynamic axis ranges. Axes sync with compare maps.
• Side-by-side synced maps (Compare View) — two maps driven by scatter plot X/Y axes, with synchronized pan/zoom and cross-highlight on hover
• Threshold highlighter — slider (0–100%) to highlight seats above or below a threshold by dimming non-matching areas
• Division filter — click a division in the breakdown to zoom and isolate it
• Neighboring seats — select a constituency to see how adjacent seats voted, with click-to-navigate
• Distribution histogram — shows how seats are distributed across the current stat's range
• Share View — generate a shareable URL that encodes the current stat, view, zoom, selected seat, and compare state. Copiable link field with one-click copy.
• Screenshot — export current map view or compare view as high-res PNG
• Mini map — inset overview map showing current viewport position
• Reset zoom — button to return to full country view
• Keyboard shortcuts — Escape to deselect, click map background to deselect

Repository structure

├── config/                 # Configuration files (coalitions.json)
├── data/                   # Input datasets (GeoJSON boundaries, saved HTML)
├── scripts/                # Python pipeline: scraping, merging, map generation
├── site_src/               # Website source (index.html)
├── site/                   # Deployed site (copied from site_src)
├── result_from_source/     # Raw scraped results from TBS News & Daily Star
├── vote_count_combined/    # Merged election results from both sources
├── official_result/        # Official Election Commission results
├── census_from_sid/        # 2022 Census data (BBS PDF extraction)
├── humdata_pop_stats/      # Subnational population statistics (HumData COD-PS)
├── unicef/                 # MICS6 district-level indicators (UNICEF)
├── poverty_data/           # District/upazila poverty statistics
├── pipeline/               # Data pipeline scripts (scraping, combining)
├── .github/workflows/      # GitHub Actions for auto-deployment
└── requirements.txt

Quick start

1. Install dependencies:

   pip install -r requirements.txt

2. Scrape vote data. The scrape_votes.py script downloads the latest TBS election page. If that fails it falls back to the saved HTML in data/tbs_election_2026.html.

   python scripts/scrape_votes.py \
     --config config/coalitions.json \
     --url https://www.tbsnews.net/election-2026 \
     --local_html data/tbs_election_2026.html \
     --output_seat_results results/seat_results.csv \
     --output_party_votes results/party_totals.csv \
     --output_party_by_seat results/party_by_seat.csv \
     --save_html

   This creates three CSV files in results/:

   • seat_results.csv – one row per constituency with vote shares for each coalition
   • party_totals.csv – national vote totals per party
   • party_by_seat.csv – vote totals per party per seat

3. Preview the site locally:

   python scripts/serve.py --dir site --port 8000

   Open http://localhost:8000 and select different statistics from the dropdown.

4. Deploy to GitHub Pages. Push to main and the GitHub Actions workflow will automatically publish the site/ directory.

Customising coalitions

Coalition definitions live in config/coalitions.json. Each entry has a unique key, display name, keyword list for matching candidate party names, and a colour gradient. Keywords are case-insensitive substring matches.

Data sources

• Election results – The Business Standard election portal and official Election Commission data
• Constituency boundaries – GeoReferenced Electoral Districts (GRED) dataset for Bangladesh's 2008 boundaries, also used by the BBC Bangla election map
• Coalition membership – based on contemporary reporting and official party announcements

Data repository

Folder	Contents
result_from_source/	Raw scraped results from TBS News and Daily Star
vote_count_combined/	Merged results from both sources
official_result/	Official Election Commission results
census_from_sid/	2022 Census data (64 districts, 8 divisions)
humdata_pop_stats/	Subnational population statistics (HumData)
unicef/	MICS6 district-level indicators
poverty_data/	Poverty statistics by district/upazila

Each folder contains its own README with column descriptions.

Credits

• Project author – S M A Nahian (https://smanahian.com)
• Vote data – The Business Standard election portal
• Boundary data – GRED / BBC Bangla
• Tooling support – Claude Code, ChatGPT agent, and GitHub Copilot

License

This project is released under the MIT License. See LICENSE for details.