# -*- coding: utf-8 -*- """ *************************************************************************** ZonalStatistics.py --------------------- Date : August 2013 Copyright : (C) 2013 by Alexander Bruy Email : alexander dot bruy at gmail dot com *************************************************************************** * * * This program is free software; you can redistribute it and/or modify * * it under the terms of the GNU General Public License as published by * * the Free Software Foundation; either version 2 of the License, or * * (at your option) any later version. * * * *************************************************************************** """ from builtins import str __author__ = 'Alexander Bruy' __date__ = 'August 2013' __copyright__ = '(C) 2013, Alexander Bruy' # This will get replaced with a git SHA1 when you do a git archive __revision__ = '$Format:%H$' import numpy try: from scipy.stats.mstats import mode hasSciPy = True except: hasSciPy = False from osgeo import gdal, ogr, osr from qgis.core import (QgsApplication, QgsRectangle, QgsGeometry, QgsFeature, QgsProcessingUtils) from processing.algs.qgis import QgisAlgorithm from processing.core.parameters import ParameterVector from processing.core.parameters import ParameterRaster from processing.core.parameters import ParameterString from processing.core.parameters import ParameterNumber from processing.core.parameters import ParameterBoolean from processing.core.outputs import OutputVector from processing.tools.raster import mapToPixel from processing.tools import dataobjects, vector class ZonalStatistics(QgisAlgorithm): INPUT_RASTER = 'INPUT_RASTER' RASTER_BAND = 'RASTER_BAND' INPUT_VECTOR = 'INPUT_VECTOR' COLUMN_PREFIX = 'COLUMN_PREFIX' GLOBAL_EXTENT = 'GLOBAL_EXTENT' OUTPUT_LAYER = 'OUTPUT_LAYER' def icon(self): return QgsApplication.getThemeIcon("/providerQgis.svg") def svgIconPath(self): return QgsApplication.iconPath("providerQgis.svg") def group(self): return self.tr('Raster tools') def __init__(self): super().__init__() self.addParameter(ParameterRaster(self.INPUT_RASTER, self.tr('Raster layer'))) self.addParameter(ParameterNumber(self.RASTER_BAND, self.tr('Raster band'), 1, 999, 1)) self.addParameter(ParameterVector(self.INPUT_VECTOR, self.tr('Vector layer containing zones'), [dataobjects.TYPE_VECTOR_POLYGON])) self.addParameter(ParameterString(self.COLUMN_PREFIX, self.tr('Output column prefix'), '_')) self.addParameter(ParameterBoolean(self.GLOBAL_EXTENT, self.tr('Load whole raster in memory'))) self.addOutput(OutputVector(self.OUTPUT_LAYER, self.tr('Zonal statistics'), datatype=[dataobjects.TYPE_VECTOR_POLYGON])) def name(self): return 'zonalstatistics' def displayName(self): return self.tr('Zonal Statistics') def processAlgorithm(self, parameters, context, feedback): """ Based on code by Matthew Perry https://gist.github.com/perrygeo/5667173 :param parameters: :param context: """ layer = QgsProcessingUtils.mapLayerFromString(self.getParameterValue(self.INPUT_VECTOR), context) rasterPath = str(self.getParameterValue(self.INPUT_RASTER)) bandNumber = self.getParameterValue(self.RASTER_BAND) columnPrefix = self.getParameterValue(self.COLUMN_PREFIX) useGlobalExtent = self.getParameterValue(self.GLOBAL_EXTENT) rasterDS = gdal.Open(rasterPath, gdal.GA_ReadOnly) geoTransform = rasterDS.GetGeoTransform() rasterBand = rasterDS.GetRasterBand(bandNumber) noData = rasterBand.GetNoDataValue() cellXSize = abs(geoTransform[1]) cellYSize = abs(geoTransform[5]) rasterXSize = rasterDS.RasterXSize rasterYSize = rasterDS.RasterYSize rasterBBox = QgsRectangle(geoTransform[0], geoTransform[3] - cellYSize * rasterYSize, geoTransform[0] + cellXSize * rasterXSize, geoTransform[3]) rasterGeom = QgsGeometry.fromRect(rasterBBox) crs = osr.SpatialReference() crs.ImportFromProj4(str(layer.crs().toProj4())) if useGlobalExtent: xMin = rasterBBox.xMinimum() xMax = rasterBBox.xMaximum() yMin = rasterBBox.yMinimum() yMax = rasterBBox.yMaximum() (startColumn, startRow) = mapToPixel(xMin, yMax, geoTransform) (endColumn, endRow) = mapToPixel(xMax, yMin, geoTransform) width = endColumn - startColumn height = endRow - startRow srcOffset = (startColumn, startRow, width, height) srcArray = rasterBand.ReadAsArray(*srcOffset) srcArray = srcArray * rasterBand.GetScale() + rasterBand.GetOffset() newGeoTransform = ( geoTransform[0] + srcOffset[0] * geoTransform[1], geoTransform[1], 0.0, geoTransform[3] + srcOffset[1] * geoTransform[5], 0.0, geoTransform[5], ) memVectorDriver = ogr.GetDriverByName('Memory') memRasterDriver = gdal.GetDriverByName('MEM') fields = layer.fields() (idxMin, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'min', 21, 6) (idxMax, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'max', 21, 6) (idxSum, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'sum', 21, 6) (idxCount, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'count', 21, 6) (idxMean, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'mean', 21, 6) (idxStd, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'std', 21, 6) (idxUnique, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'unique', 21, 6) (idxRange, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'range', 21, 6) (idxVar, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'var', 21, 6) (idxMedian, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'median', 21, 6) if hasSciPy: (idxMode, fields) = vector.findOrCreateField(layer, fields, columnPrefix + 'mode', 21, 6) writer = self.getOutputFromName(self.OUTPUT_LAYER).getVectorWriter(fields, layer.wkbType(), layer.crs(), context) outFeat = QgsFeature() outFeat.initAttributes(len(fields)) outFeat.setFields(fields) features = QgsProcessingUtils.getFeatures(layer, context) total = 100.0 / QgsProcessingUtils.featureCount(layer, context) for current, f in enumerate(features): geom = f.geometry() intersectedGeom = rasterGeom.intersection(geom) ogrGeom = ogr.CreateGeometryFromWkt(intersectedGeom.exportToWkt()) if not useGlobalExtent: bbox = intersectedGeom.boundingBox() xMin = bbox.xMinimum() xMax = bbox.xMaximum() yMin = bbox.yMinimum() yMax = bbox.yMaximum() (startColumn, startRow) = mapToPixel(xMin, yMax, geoTransform) (endColumn, endRow) = mapToPixel(xMax, yMin, geoTransform) width = endColumn - startColumn height = endRow - startRow if width == 0 or height == 0: continue srcOffset = (startColumn, startRow, width, height) srcArray = rasterBand.ReadAsArray(*srcOffset) srcArray = srcArray * rasterBand.GetScale() + rasterBand.GetOffset() newGeoTransform = ( geoTransform[0] + srcOffset[0] * geoTransform[1], geoTransform[1], 0.0, geoTransform[3] + srcOffset[1] * geoTransform[5], 0.0, geoTransform[5], ) # Create a temporary vector layer in memory memVDS = memVectorDriver.CreateDataSource('out') memLayer = memVDS.CreateLayer('poly', crs, ogr.wkbPolygon) ft = ogr.Feature(memLayer.GetLayerDefn()) ft.SetGeometry(ogrGeom) memLayer.CreateFeature(ft) ft.Destroy() # Rasterize it rasterizedDS = memRasterDriver.Create('', srcOffset[2], srcOffset[3], 1, gdal.GDT_Byte) rasterizedDS.SetGeoTransform(newGeoTransform) gdal.RasterizeLayer(rasterizedDS, [1], memLayer, burn_values=[1]) rasterizedArray = rasterizedDS.ReadAsArray() srcArray = numpy.nan_to_num(srcArray) masked = numpy.ma.MaskedArray(srcArray, mask=numpy.logical_or(srcArray == noData, numpy.logical_not(rasterizedArray))) outFeat.setGeometry(geom) attrs = f.attributes() v = float(masked.min()) attrs.insert(idxMin, None if numpy.isnan(v) else v) v = float(masked.max()) attrs.insert(idxMax, None if numpy.isnan(v) else v) v = float(masked.sum()) attrs.insert(idxSum, None if numpy.isnan(v) else v) attrs.insert(idxCount, int(masked.count())) v = float(masked.mean()) attrs.insert(idxMean, None if numpy.isnan(v) else v) v = float(masked.std()) attrs.insert(idxStd, None if numpy.isnan(v) else v) attrs.insert(idxUnique, numpy.unique(masked.compressed()).size) v = float(masked.max()) - float(masked.min()) attrs.insert(idxRange, None if numpy.isnan(v) else v) v = float(masked.var()) attrs.insert(idxVar, None if numpy.isnan(v) else v) v = float(numpy.ma.median(masked)) attrs.insert(idxMedian, None if numpy.isnan(v) else v) if hasSciPy: attrs.insert(idxMode, float(mode(masked, axis=None)[0][0])) outFeat.setAttributes(attrs) writer.addFeature(outFeat) memVDS = None rasterizedDS = None feedback.setProgress(int(current * total)) rasterDS = None del writer