14 synthetic datasets, 72,860 rows in total, each built around a specific analytical trap. Download the CSV, or open one in a browser tool without downloading anything.
14 of 14 shown
Eight thousand city ride-hail trips with fares, distances, surge multipliers and tips. Heavy right skew on fare, a rush-hour bimodal duration distribution, and a handful of impossible trips left in on purpose.
Seven thousand e-commerce order lines across four channels and six categories, with a real Q4 seasonality bump, a returns flag, and discounting that varies by channel.
Twenty-five hundred B2B subscription accounts with signup dates, plans, seat counts, MRR and churn dates. Built for cohort retention, survival analysis, and expansion-revenue maths.
Eight thousand readings from twelve machines across three sites, at ten-minute intervals, with drift, seasonal temperature, injected anomalies, and gaps where a gateway dropped offline.
Eight thousand domestic flights across six carriers and fourteen airports, with departure and arrival delay, cancellations, and a delay distribution that is long-tailed rather than normal.
A synthetic two-arm trial: 900 subjects, baseline and week-12 scores, adverse events, and a real but modest treatment effect that is easy to overstate if you analyse it badly.
Daily pollutant readings for four cities over nearly four years, with strong seasonality, weekday effects, wildfire episodes, and genuine sensor outages.
Thirty-five hundred data job ads with titles, seniority, location, remote policy, salary bands, and a pipe-delimited skills column that needs parsing before it is useful.
Six thousand users randomised across a control and two variants, with conversion, revenue, device, and a novelty effect that fades over the first week.
A year of hourly electricity demand, renewable generation, temperature and wholesale price for one balancing area — including the hours when price goes negative.
Nine thousand user ratings over 260 films, with genres, release years, and the long-tail popularity distribution that makes naive recommenders look better than they are.
A tidy panel of thirty countries over twenty-four years with GDP per capita, life expectancy, population, CO₂ and internet penetration — small enough to hold in your head, rich enough for real modelling.
Four thousand support tickets with free-text subjects, priorities, queue routing, first-response and resolution times, and SLA breaches concentrated in exactly the places you would not guess.
A thousand customer records broken in every way real data is broken: five date formats, inconsistent casing, padded whitespace, near-duplicate rows, mixed units, sentinel nulls, and encoding damage.