Skip to content

Features

Two analysis modes

The plugin ships as two Processing algorithms and two dialog tabs, so the analysis kind is always an explicit choice.

Continuous Categorical
For measured values (elevation, NDVI) class codes (land cover, soil)
Statistics count, nodata, min, max, sum, mean, stddev majority, minority, variety, count, nodata
Layouts Long, Wide Summary, Class breakdown, Class counts

See Continuous & Categorical for the full details and examples.

Multiband in one pass

Every selected band is aggregated in a single tiled sweep of the raster. Band selections accept all, ranges like 1-8, and lists like 1,3,7.

Output that's easy to work with

  • Unique fid

    Every row starts with a 1-based fid — a stable primary key for joins and database imports, since polygon_id repeats across bands and classes.

  • Decimal places

    Keep full precision, or fix numeric cells to 0–15 decimals for clean, spreadsheet-friendly numbers.

  • Band names

    Columns keep the raster's own band names, with an optional common prefix and suffix (e.g. 2020_ndvi).

  • Atomic CSV

    Output is written to a temporary file and renamed on success — a cancelled or failed run never leaves a partial CSV.

Exclude values (extra nodata)

Type a comma-separated list of pixel values into Exclude values to drop them from the calculation, just like nodata — for example 0 for out-of-boundary pixels in a categorical raster, or 0, 255. They apply to every band, on top of the raster's declared nodata.

Histogram preview

The Histogram… button previews a band's distribution in a popup — binned values (continuous) or one bar per class (categorical). It describes the whole raster (read decimated for speed) and is a quick way to decide which values to exclude. Drawn with Qt directly, so it needs no extra Python packages.

Coordinate systems

The raster's CRS is the analysis CRS and the raster is never resampled. Zones are transformed into the raster CRS automatically. Optional Raster CRS and Zones CRS overrides assign a CRS to a layer whose declared CRS is missing or wrong — they reinterpret coordinates, they do not reproject data.

Robust and fast

  • Overlapping polygons — shared pixels are counted for every overlapping zone (overlaps are rasterized in separate passes).
  • Multipart features keep one zone ID; holes are respected.
  • Per-band nodata is honored, and the population standard deviation uses a numerically stable parallel-variance merge.
  • Bounded memory via configurable tile size and band chunk settings, plus pixel-centre or all-touched inclusion.

Everywhere in QGIS

Available from the Raster menu, the toolbar, the graphical dialog, the Processing Toolbox, and Model Builder — the dialog and Processing runs share the same engine.