Percentile

Summary

Creates a raster object in which each pixel contains the specified percentile value across multiple rasters.

Discussion

The referenced raster dataset for the raster object is temporary. To make it permanent, you can call the raster object's save method.

Syntax

Percentile (rasters, {percentile_values}, {percentile_interpolation_type}, {extent_type}, {cellsize_type}, {ignore_nodata}, {process_as_multiband})

Parameter	Explanation	Data Type
rasters [rasters,...]	The list of input rasters.	Raster
percentile_values	The percentile value to calculate. (The default value is 90)	Double
percentile_interpolation_type	The interpolation method used when the specified percentile value lies between two input cell values. AUTO_DETECT—If the input raster collection is integer type, the NEAREST method is used. If the input raster collection is floating point type, the LINEAR method is used. NEAREST—The nearest available value to the specified percentile is used. The output pixel type will match the input. LINEAR—The weighted average of the two surrounding percentile values is used. The output pixel type is floating point. (The default value is AUTO_DETECT)	String
extent_type	The method that will be used to compute the extent of the output when the input rasters have different extents. FirstOf—The output extent will be defined using the extent of the first input raster. LastOf—The output extent will be defined using the extent of the last input raster. IntersectionOf—The output extent will be defined as the intersecting area of the input rasters. UnionOf—The output extent will be defined as the total extent of the input rasters. (The default value is FirstOf)	String
cellsize_type	The method that will be used to compute the pixel size of the output when the input rasters have different pixel sizes. FirstOf—The output pixel size will be defined using the pixel size of the first input raster. LastOf—The output pixel size will be defined using the pixel size of the last input raster. MaxOf—The output pixel size will be defined using the maximum pixel size of the input rasters. MeanOf—The output pixel size will be defined using the mean of both input rasters. MinOf—The output pixel size will be defined using the minimum pixel size of the input rasters. (The default value is FirstOf)	String
ignore_nodata	Specifies whether NoData values will be ignored in the calculation. True—The analysis will include all valid pixels in the input rasters and ignore any NoData pixels. False—The analysis will result in NoData if there are any NoData values for the pixels in the input rasters. (The default value is False)	Boolean
process_as_multiband	Specifies how the bands of the input rasters will be processed. True—Each multiband raster will be processed as a multiband raster. The operation will be performed for each band from one input using the corresponding band number from other inputs. False—Each band from a multiband raster input will be processed separately as a single-band raster. (The default value is False)	Boolean

Return Value

Data Type	Explanation
Raster	The output raster.

Code sample

Percentile example

Computes a raster object that contains the 90th percentile value across three multiband rasters.

from arcpy.ia import * 

rc = RasterCollection([landsat_1, landsat_2, landsat_3]) 

percentile_raster = Percentile(rc, percentile_value = 90, percentile_interpolation_type = "AUTO_DETECT", extent_type = "UnionOf", cellsize_type = "MinOf", 
		ignore_nodata = True, process_as_multiband = True)