Correlation Matrix for Sensor Channels — Pearson and Spearman
Load a multi-channel export and see how every channel relates to every other. Each cell carries its coefficient and the number of readings the two channels actually share, so a −0.4 from nine hundred readings never looks like a −0.4 from nine — and switching between Pearson and Spearman tells you whether a weak-looking relationship is weak or just curved.
Your data
Drop your logger CSV or Excel file here
CSV, TSV and .xlsx workbooks. HOBOware and Postlad exports are recognised on sight, and most other files are read anyway. It stays on this machine — the workbook is even unzipped here, and nothing is uploaded.
Synthetic seven-day record from a seven-channel weather station at fifteen-minute intervals, written as a Postlad stream export. It contains a wet spell on the fourth afternoon, a six-hour humidity-sensor dropout on the third day, and a leaf-wetness sensor installed on the fourth morning — so roughly half of that channel’s cells are marked as low-overlap. Not measured data.
Saved projects
0 of 1Projects are stored in this browser only — never uploaded, and gone if you clear site data. Download one as a file to keep it, move it to another machine, or send it to a colleague.
Click any title, axis label, legend or note on the figure to reword it — Enter keeps it, Escape cancels, and an empty box hides the line. The numbers are not editable.
SVG is vector — scale it to any size in Illustrator, Inkscape or LaTeX without losing quality, and edit the type. Use it unless a journal insists on raster.
Options
Statistic
Pearson measures straight-line association. Spearman ranks the readings first, so it catches any monotone relationship and is unmoved by outliers.
Figure
The matrix is symmetric, so the lower triangle holds every distinct pair exactly once.
More options — statistic, figure
Statistic
Averaging both series before correlating them removes independent noise from each and raises the coefficient. Legitimate — but the figure will say it was done.
A pair with fewer readings in common than this is left blank rather than given a coefficient.
Two channels that only overlap for part of the record are correlated over that part. The mark says so. Set to 0 to switch it off.
Figure
Purely a reading aid. A coefficient does not become meaningful by crossing a round number.
Grouping puts strongly-related channels next to each other, so blocks of related sensors become visible.
All of that was worked out in this browser tab. Your file was not uploaded, and no request goes out while you work — open the network panel and watch, if you like.
How this tool’s arithmetic is checked — the published worked examples this tool reproduces, printed beside what it computes, with the test files that assert them.