Log axes with 10ⁿ ticks. Stats computed in the engine. Neon on black. Same JSON, same picture, forever.
sci-core is the scientific module on the same engine as the rest of SlickFast. You write a spec. You get SVG. Scientists check the decade grid first — labels in power notation, minor lines bunching toward the top of each decade. That furniture is the product.
This Page is the showcase. Add `.json` to the URL for the twin.
E. coli, 120 → 71,000 CFU/mL. Linear hides the first hour. Log-y makes every decade a floor.
Two compounds. Error bars in the spec. The engine draws whiskers on the same 10ⁿ furniture.
Raw times and events go in. The estimator, censor ticks, and median are computed at render.
Fold-changes and raw p-values in. The engine computes −log₁₀(p), draws thresholds, colors up / down / not significant.
A chromatogram-shaped lineplot. `peaks: true` — the engine finds them.
Normality and method agreement from raw series. Stats stay in the engine.
lineplot (linear + log-y) · errorbar (log–log) · survival · volcano · peaks · qq · blandaltman.
Also in sci-core: scatter, barplot, trend fits, error bands, residual panels, reference lines, base-2 axes.
Same spec, same bytes. Twin at `.json`. The hash is on the page.