Every chart Twin Vector Labs publishes. We build these for the fun of it — they are not a product, and nothing here is for sale. The data underneath is the real thing: each series comes from the organisation that publishes it and stays current on that publisher’s schedule.
A 0–100 gauge of supply-chain stress, blending four public signals: news reports of disruption from GDELT, diesel prices from the EIA, import values from the Census Bureau, and truck-freight prices from the BLS. Each input is scored against its own recent history, so 50 is an unremarkable month and the distance from it is what carries meaning.
Annual revenue, indexed to a common start, for commercial construction and the industries that feed it — steel, cement and aggregates, construction machinery, and HVAC. The supplier list is derived from the Bureau of Economic Analysis input-output accounts, and the revenue is what the public companies in each industry reported to the SEC. It shows which suppliers actually grew as the data-center build-out arrived.
Rainfall over the last 30 days in each state, measured against that state's own average for the same weeks across 2021–2025, with the major corn, soybean and wheat states emphasised. Precipitation comes from NOAA's GHCN-Daily station network; a grains producer-price line from the BLS runs alongside it for context.
Ten critical minerals scored on three exposures at once: how concentrated their production is in a few countries, how far their price has moved, and how much tariff pressure sits on them. Production and reserves come from the US Geological Survey, prices from the International Monetary Fund, and import exposure from Census trade records.
Weekly preprint volume in a research area — arXiv, joined by bioRxiv and medRxiv for the life sciences — set against federal research and development obligations to the matching industry cohort. The two series are compared at a range of offsets, and the offset that lines them up best is the headline: how far behind the published work the funding runs.
Twelve commodity groups from the Bureau of Labor Statistics Producer Price Index, each with a 24-month sparkline. The index measures what domestic producers are paid for their output, so the wall reads as a survey of where price pressure sits across energy, metals and agriculture.
Fuel prices set against how the news talks about oil. The price side is the Bureau of Labor Statistics Producer Price Index for fuels and related products and power; the tone side is the sentiment of oil-related news coverage from GDELT, averaged by month. Both axes are fitted to their own range, so what the chart compares is the shape of the two lines rather than their levels — the interesting months are the ones where they part.
How the twelve Producer Price Index groups on the Commodities Wall have moved relative to one another, as the pairwise correlation of their monthly changes over three years. Groups that rise and fall together cluster into the familiar macro categories; the cells that run the other way mark the pairs that have historically moved in opposition.
