What this site does. You bring a table of numbers, it gives you
numbers back. This page slides one signal along another and computes the
correlation coefficient r at every shift: the shift with the highest r is the time
delay between them, and a signal against itself shows its period. It runs entirely
in your browser — nothing is uploaded, nothing is stored. If you want the reasoning
rather than the answer, the cross-correlation
article starts from Pearson's r and explains every step.
Correlation needs evenly spaced samples, and t must be time in seconds: a lag is a
fixed number of samples. Remove a trend first - a trend correlates with itself at
every lag.
- Paste your data into the Input grid — straight from a
spreadsheet, or as CSV/TSV text — or import a CSV/TSV file.
Column t (s), then up to three y columns. Up to 5,000 rows.
- Rows are sorted by t. A row with no number in t is skipped; a blank y cell is
skipped for that column only - which leaves a gap, so that column is refused
as unevenly spaced.
- Pick a mode. Cross compares every column with y1, the
reference: a positive lag means the column is later than y1. Auto compares
each column with itself, to find a period.
- Pick the normalization. Biased divides by the full record
length, so r shrinks at large lags; unbiased divides by the number of pairs at
each lag, and grows noisy at large lags.
- Real time range — try settings live on the max lag box.
- ▲▼ buttons or the mouse wheel over the
box: the value moves and the result is recomputed at once - no need to
press correlate.
- Digit wheel: over the box, the wheel steps the digit just
left of the mouse pointer.
- Limits: the max lag stays above 0; a typed value
outside is set to the nearest limit.
- Typing or the arrow keys change the box only; press
correlate to recompute.
- Press correlate. The upper chart shows your data; the lower
chart shows r against lag, with dashed lines at ±1.96/√N - the range unrelated
white noise stays inside 95% of the time. The status line gives the peak lag
(cross) or the period (auto).
- Read the result. The Output grid lists the lag in seconds,
then r for each pair or column. Export saves it as a CSV/TSV
file.