Output & Options¶
The fid column¶
Every output row — in every layout — begins with a unique 1-based fid, assigned in write order. It's the stable identity of each row: polygon_id repeats across bands (and across classes in the breakdown), so fid is the real primary key for joins, database imports, or referencing a specific row.
Decimal places¶
Control the decimals of numeric cells:
- Full precision (the default) keeps every digit.
- Set 0–15 to fix the decimals — only floating-point cells (statistics, fraction) are formatted; integers stay clean.
Numbers look like text in Excel?
That's usually a locale mismatch — the CSV uses a . decimal but your Excel expects a ,. Import via Data → From Text/CSV and set the decimal separator, or reduce the output to a few decimal places.
Band names¶
Output columns keep the raster's own band descriptions (falling back to band_001, band_002, … when unnamed). Add a common prefix and suffix to label a run of files — the raster's bands are never renamed, only the CSV labels.
| Band name | Prefix | Suffix | Wide column (sanitised) |
|---|---|---|---|
| NDVI | 2020_ | — | 2020_ndvi_mean |
| Surface Temp | — | — | surface_temp_mean |
| band 1 | — | _dry | band_1_dry_mean |
Wide-format headers are lowercased, with spaces and punctuation turned into underscores and any collisions numbered.
Exclude values (extra nodata)¶
Type a comma-separated list into Exclude values to drop those pixel values from the calculation, just like nodata:
- Categorical rasters often use
0for out-of-boundary pixels — exclude it so it doesn't dominate the class counts. - Multiple sentinels work too, e.g.
0, 255.
They apply to every band, on top of the raster's declared nodata. Use the histogram first to spot which value to exclude.
Histogram preview¶
The Histogram… button opens a quick preview of a band's distribution:
- Pick any band, and toggle Continuous (binned values, with an adjustable bin count) or Categorical (one bar per class). It opens on your active tab's mode.
- It describes the whole raster (it ignores the polygons) and reads the band decimated for speed, so even large rasters preview promptly.
- Summary stats sit beneath the chart: min / max / mean / sd for continuous, or the class count and majority for categorical.
Advanced settings¶
| Setting | What it does |
|---|---|
| Tile size | Raster window read at a time. Larger = fewer reads, more memory. |
| Band chunk | Bands held in memory per pass. Lower it for many-band rasters on tight memory. |
| All touched | Include every pixel a polygon touches, not just pixel centres (slower). |
Atomic output¶
The CSV is written to a temporary file and renamed on success, so a cancelled or failed run never publishes a partial result. Output is UTF-8.