Validation¶

The plugin includes comprehensive validation to ensure your inputs are compatible and will produce meaningful results. Always validate before running an analysis.

Running Validation¶

Configure your inputs in the Inputs tab
Click the Validate Inputs button
Switch to the Validation tab to see results

Validation Results¶

Each check displays one of three status indicators:

Status	Meaning
PASS	Check passed - no issues
WARN	Warning - analysis can proceed but review recommended
FAIL	Failed - must be fixed before analysis

FAIL Items

You cannot run the analysis if any check shows FAIL. Fix the underlying issue and re-validate.

Validation Checks¶

CRS Consistency¶

Check: All rasters use the same Coordinate Reference System.

PASS: All rasters have identical CRS.

FAIL: Rasters have different CRS values.

How to fix:

Reproject rasters to a common CRS using QGIS raster tools
Use Raster > Projections > Warp (Reproject)

Reprojection Warning

When reprojecting categorical rasters, always use Nearest Neighbor resampling to preserve class values.

Pixel Size Consistency¶

Check: All rasters have identical pixel dimensions.

PASS: All rasters have the same X and Y pixel size.

FAIL: Pixel sizes differ between rasters.

How to fix:

Resample rasters to a common resolution
Use Raster > Projections > Warp (Reproject) with target resolution

Extent Consistency¶

Check: All rasters cover the same geographic extent.

PASS: All rasters have identical bounding boxes.

FAIL: Extents differ between rasters.

How to fix:

Clip rasters to a common extent
Use Raster > Extraction > Clip Raster by Extent

Dimensions Consistency¶

Check: All rasters have the same width and height in pixels.

PASS: All rasters have identical dimensions.

FAIL: Raster dimensions differ.

How to fix:

Usually resolved by fixing extent and pixel size issues
Ensure all rasters were produced with the same grid specification

Grid Alignment¶

Check: Pixel grids are aligned (same origin coordinates).

PASS: All rasters have aligned pixel grids.

FAIL: Pixel origins are offset between rasters.

How to fix:

Align rasters using Processing Toolbox > GDAL > Raster projections > Warp
Set a specific target extent that aligns with your grid

Why Grid Alignment Matters

Even with matching extents and pixel sizes, rasters can be offset by fractions of a pixel. This causes each pixel to represent slightly different areas, making per-pixel comparisons invalid.

Data Type Check¶

Check: Rasters contain integer data types.

PASS: All rasters use integer types (Byte, Int16, Int32, etc.).

WARN: One or more rasters use floating-point types.

How to fix (if needed):

Convert float rasters to integer
Use Raster > Conversion > Translate and set output type

Float Warning

Float rasters can still be processed, but they're unusual for categorical LULC data. The warning ensures you're aware of the data type.

NoData Presence¶

Check: NoData values are defined in raster metadata.

PASS: All rasters have NoData values defined.

WARN: Some rasters lack NoData definitions.

How to fix:

Use the Override option in the Inputs tab to specify a NoData value
Or set NoData in the source raster using GDAL tools

NoData Consistency¶

Check: All rasters use the same NoData value.

PASS: NoData values are consistent.

WARN: Different NoData values are used.

How to fix:

Use the Override option to specify a single NoData value for all rasters

Value Range Detection¶

Status: Always INFO (informational only)

Purpose: Reports the range of class values found in each raster.

Example output: Raster values: 0-15

Use This Information

Verify the range matches your expected class scheme
Identify unexpected values that might indicate data issues
Ensure all rasters use the same classification scheme

Unique Class Count¶

Status: Always INFO (informational only)

Purpose: Reports the number of unique class values in each raster.

Example output: 12 unique classes detected

Class Count Limit

Detection is capped at 1024 classes for performance. If you have more than 1024 classes, only the first 1024 are counted.

AOI Coverage¶

Status: Always INFO (informational only, only shown when AOI is selected)

Purpose: Reports what percentage of each raster is covered by the AOI.

Example output: AOI covers 45.2% of raster extent

Coverage Considerations

Very low coverage might indicate AOI/raster misalignment
100% coverage means AOI fully contains the raster
Coverage > 100% can occur if AOI extends beyond raster bounds

Validation Best Practices¶

Before Analysis¶

Always validate - Never skip validation
Fix all FAIL items - Analysis cannot proceed with failures
Review all WARN items - Understand why warnings occurred
Check INFO items - Verify values match expectations

Common Validation Scenarios¶

All PASSSome WARNAny FAIL

Ready to run analysis. Proceed to the Options tab to configure outputs.

Analysis can proceed, but review warnings:

Float data type: Verify this is expected
NoData inconsistency: Consider using Override
Missing NoData: Specify a value if needed

Must fix before proceeding:

Note which checks failed
Use QGIS tools to fix source data
Reload rasters if modified externally
Re-run validation

Troubleshooting Validation¶

"CRS mismatch" but CRS looks the same¶

Different CRS definitions can represent the same system:

EPSG:4326 vs WGS 84 (same but different string)
Check the actual EPSG codes
Reproject all to a single well-defined CRS

"Extent mismatch" by tiny amounts¶

Floating-point precision can cause tiny differences:

Clip all rasters to a common extent
Use the same bounding box coordinates for all

Validation takes too long¶

For very large rasters:

Validation scans class values, which takes time
Consider working with a subset for testing
The actual analysis uses efficient block processing

Next Steps¶

Once all checks pass:

Review Configuration options
Proceed to Running Analysis