📘 How to use this — copy and paste CSV, TSV, or spreadsheet data

What this site does. You bring a table of numbers, it gives you numbers back. This page gives every point a score - how far it lies from the center of its column, counted in spreads - and flags the points whose score passes a threshold. It runs entirely in your browser — nothing is uploaded, nothing is stored. If you want the reasoning rather than the answer, the outlier detection article explains every method on a worked example.

A flag is a question, not a verdict: look at each flagged point before you remove it. A long tail (response times, sizes, concentrations) is flagged by every method here - take the logarithm first.

  1. Paste your data into the Input grid — straight from a spreadsheet, or as CSV/TSV text — or import a CSV/TSV file. Column t (time, or a row number), then up to three y columns. Up to 5,000 rows.
  2. Rows are sorted by t. A row with no number in t is skipped; a blank y cell is skipped for that column only. Each column is judged on its own, and needs at least 3 numbers.
  3. Pick a method.
    • z-score: distance from the mean in standard deviations. One large outlier can hide the others.
    • modified z-score: distance from the median in MADs, scaled like a z-score. Not fooled by the outliers themselves.
    • IQR: distance beyond the middle half of the data, in interquartile ranges - Tukey's fences, as on a box plot.
    • Hampel: the modified z-score within a window of neighbors around each point. For drifting signals.
  4. Set the threshold. Picking a method fills in its usual value: 3 for the z-score, 3.5 for the modified z-score, 1.5 for the IQR, 3 for Hampel. A larger value flags fewer points.
  5. Set the window (Hampel only): the number of points on each side, so the window holds 2 × window + 1 points. It must be longer than a glitch and shorter than the changes in your signal.
  6. Real time range — try settings live on the threshold and window boxes.
    • ▲▼ buttons or the mouse wheel over a box: the value moves and the result is recomputed at once - no need to press detect.
    • Digit wheel: over the threshold box, the wheel steps the digit just left of the mouse pointer.
    • Limits: threshold 0.1 to 20, window 1 to 50; a typed value outside is set to the nearest limit.
    • Typing or the arrow keys change the box only; press detect to recompute.
  7. Press detect. The upper chart shows your data with the threshold as dotted upper and lower lines in a lighter shade of each column's color, in the data's own units: a point beyond them is an outlier. For Hampel the lines follow the window. The lower chart shows each column's score against t, with the outliers marked in a darker shade. The status line lists the flagged t for each column that has any.
  8. Read the result. The Output grid lists t, then for each column its score and a flag (1 = outlier, 0 = not). Export saves it as a CSV/TSV file.

Input

Output

Signal

Score