“Show for last four years different us senators positions on different questions with clustering and visualisation”
A curated, source-extracted sample of recorded Senate roll-call positions, 2023 through September 1, 2026. Twelve consequential questions were targeted; after conservative cleaning, the sources yielded usable recorded positions for 58 senators on three of them — the 2024 border-security cloture (S.4361), the 2024 contraception-access cloture (S.4381), and the 2026 Russia–Iran sanctions passage vote (H.R.5334) — 67 senator–question positions in all. This is a selected sample, not every Senate vote and not an ideology score. Cloture votes are procedural and are not final passage.
Hover a cell for the senator, question, position and source. Most cells are blank: coverage is incomplete and blanks are missing records, never votes.
Agglomerative (complete-linkage) clustering on mismatch rate over mutually observed votes only; senators with no shared vote are never merged. Cluster names describe this selected vote pattern, not ideology.
| Senator | Party | State | Measure | Year | Question | Position | Source |
|---|
Method: 3,122 web-extracted rows about twelve targeted Senate questions (2023–2026 through 2026-09-01) were cleaned conservatively — positions normalized to Yea/Nay/Not Voting/Present, rows without a recorded position dropped, identical senator–measure duplicates merged, no conflicts found. That left 67 recorded positions for 58 senators on three measures; the other nine targeted questions (including 2023 H.J.Res.30, H.R.3746, H.R.2670; 2024 H.R.815; 2025 S.5, H.R.1, H.R.4; 2026 H.R.6644, S.2) had no extractable per-senator positions and are shown nowhere as votes. Cloture is procedural, not passage. Similarity uses only mutually observed votes; blanks are missing records. Extraction noise and partial coverage apply; 2026 is year-to-date.