Reading temp_c before trusting it in sensor-telemetry

Profiling temp_c on sensor-telemetry before anyone builds anything on top of it.

8,000 rows, no nulls, 1,414 distinct values. Machine temperature in Celsius.

The five-number version: min 32.72, p10 37.54, median 41.28, p90 45.16, max 71.74. The mean is 41.42.

The p10 to p90 band — 37.54 to 45.16 — is where ordinary rows live, and it is the pair worth quoting when somebody asks what to expect. Min and max describe the two strangest rows in the table and nothing else; they are useful for spotting impossible values and misleading for everything else.

select
  count(*)                 as rows,
  count(temp_c)            as present,
  count(*) - count(temp_c) as nulls,
  count(distinct temp_c)   as distinct_values
from sensor_telemetry;

Nothing dramatic: the mean and median are within 0.3% of each other, so an average is a fair summary. That is worth confirming rather than assuming — it is not true of most money columns.

Where this matters: a symmetric column is one you can average, threshold and chart without hedging, which makes it unusually cheap to work with. Knowing which of your columns are like this and which are not is most of knowing when to be careful.

The grain is one row per sensor reading, which is the context every one of those numbers depends on. None of them survive a change of grain, which is why "profile the column" and "profile the table" are the same job. Full schema, and the CSV, on the dataset page.