RS_MapAlgebra

Introduction: Apply a map algebra script on a raster.

Warning

RS_MapAlgebra is deprecated since v1.9.1 and will be removed in a future version. Use a Raster UDF instead, which expresses the same per-pixel calculations in Python and additionally reaches NumPy, SciPy, scikit-learn, and rasterio. See the Raster UDF page for the same NDVI example written both ways, and migrate existing scripts.

Format:

RS_MapAlgebra (raster: Raster, pixelType: String, script: String)

RS_MapAlgebra (raster: Raster, pixelType: String, script: String, noDataValue: Double)

RS_MapAlgebra(rast0: Raster, rast1: Raster, pixelType: String, script: String, noDataValue: Double)

Return type: Raster

Since: v1.5.0

RS_MapAlgebra runs a script on a raster. The script is written in a map algebra language called Jiffle. The script takes a raster as input and returns a raster of the same size as output. The script can be used to apply a map algebra expression on a raster. The input raster is named rast in the Jiffle script, and the output raster is named out.

SQL Example

Calculate the NDVI of a raster with 4 bands (R, G, B, NIR):

-- Assume that the input raster has 4 bands: R, G, B, NIR
-- rast[0] refers to the R band, rast[3] refers to the NIR band.
SELECT RS_MapAlgebra(rast, 'D', 'out = (rast[3] - rast[0]) / (rast[3] + rast[0]);') AS ndvi FROM raster_table

Output:

+--------------------+
|              raster|
+--------------------+
|GridCoverage2D["g...|
+--------------------+

Spark SQL Example for two raster input RS_MapAlgebra:

RS_MapAlgebra(rast0, rast1, 'D', 'out = rast0[0] * 0.5 + rast1[0] * 0.5;', null)

For more details and examples about RS_MapAlgebra, please refer to the Map Algebra documentation. To learn how to write map algebra script, please refer to Jiffle language summary. To express the same calculation in Python with NumPy or rasterio instead, see Raster UDFs.