These standalone tools let you explore the mathematics and politics of poverty measurement at your own pace — no path or prior reading required. Drag parameters, build point clouds, and compare how different assumptions change who counts as poor.
Compare the persistence barcodes produced when the same 30 households are described using official DWP deprivation indices vs community-defined dimensions (safety, trust, green space). The bottleneck distance between the two diagrams is the mathematical signature of the difference between being measured and being heard.
Explore how Universal Credit's 55% taper rate and work allowances determine net income at every earnings level. Toggle the housing element, enable a comparison with the pre-December 2021 63% taper, and spotlight any earnings point to see the effective marginal rate you face.
Adjust the decision threshold τ of a logistic regression welfare scoring model and observe the resulting TPR, FPR, precision, and F1. A sub-group comparison panel reveals how the same τ produces different error rates for majority (Group A) and minority (Group B) claimants — the sociotechnical mechanism of algorithmic disparity.
Apply three UK equivalisation scales — Original OECD, Modified OECD, and McClements — to the same stylised household income distribution. Poverty rates diverge purely from the formula choice. Adjust the threshold slider to explore sensitivity across the 40–75% of median range.
Watch a Vietoris-Rips filtration grow step by step: place points, then sweep the radius to see balls expand, edges form, and triangles fill. Track connected components and loops as they are born and die.
Click edges and triangles to add or remove them from a simplicial complex. Watch β₀ (connected components) and β₁ (independent loops) update in real time. Three presets explore the triangle, the torus skeleton, and the Möbius strip.
Explore how the Mapper algorithm turns a point cloud into a topological graph by adjusting cover resolution, interval overlap, filter function, and clustering threshold in real time.
An interactive visualisation of the normal distribution with draggable parameters (μ and σ) and historical overlays tracing how the bell curve has been used — and misused — to classify human populations, from Quetelet's "average man" to IQ scoring and benefit thresholds.
An interactive Vietoris-Rips filtration explorer. Place a point cloud, then sweep the filtration radius to watch connected components merge and loops appear and disappear. The persistence diagram records the birth and death of every topological feature, letting you read the shape of data directly from the scatter of points above the diagonal.
Click any point to see its ε-ball expand. Toggle between Euclidean (circle) and Manhattan (diamond) distance to see how metric choice changes neighbourhood structure. Expand the distance matrix to inspect all pairwise distances.
An interactive visualisation demonstrating that poverty thresholds are political choices, not neutral facts. Compare three measurement traditions — the Rowntree/JRF Minimum Income Standard (absolute), the Townsend/EU AROPE 60%-of-median approach (relative), and the DWP HBAI Before Housing Costs series (official UK statistics) — for any household composition.
Perturb the feature values of a near-threshold welfare claimant and watch SHAP attributions shift while the predicted score stays approximately constant. Demonstrates that SHAP explanations are properties of a local approximation algorithm — not the model's internal representations.
Explore pre-computed Vietoris-Rips persistence diagrams for four canonical point clouds. Sweep the filtration radius to watch connected components merge and loops appear in real time alongside the birth-death scatter.