The code snippets on this page need the following imports if you’re outside the pyqgis console:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
from qgis.core import ( QgsRasterLayer, QgsProject, QgsPointXY, QgsRaster, QgsRasterShader, QgsColorRampShader, QgsSingleBandPseudoColorRenderer, QgsSingleBandColorDataRenderer, QgsSingleBandGrayRenderer, ) from qgis.PyQt.QtGui import ( QColor, )
Using Raster Layers¶
A raster layer consists of one or more raster bands — referred to as single band and multi band rasters. One band represents a matrix of values. A color image (e.g. aerial photo) is a raster consisting of red, blue and green bands. Single band rasters typically represent either continuous variables (e.g. elevation) or discrete variables (e.g. land use). In some cases, a raster layer comes with a palette and the raster values refer to the colors stored in the palette.
The following code assumes
rlayer is a
rlayer = QgsProject.instance().mapLayersByName('srtm') # get the resolution of the raster in layer unit print(rlayer.width(), rlayer.height())
# get the extent of the layer as QgsRectangle print(rlayer.extent())
<QgsRectangle: 20.06856808199999875 -34.27001076999999896, 20.83945284300000012 -33.75077500700000144>
# get the extent of the layer as Strings print(rlayer.extent().toString())
20.0685680819999988,-34.2700107699999990 : 20.8394528430000001,-33.7507750070000014
# get the raster type: 0 = GrayOrUndefined (single band), 1 = Palette (single band), 2 = Multiband print(rlayer.rasterType())
# get the total band count of the raster print(rlayer.bandCount())
# get all the available metadata as a QgsLayerMetadata object print(rlayer.metadata())
<qgis._core.QgsLayerMetadata object at 0x13711d558>
When a raster layer is loaded, it gets a default renderer based on its type. It can be altered either in the layer properties or programmatically.
To query the current renderer:
<qgis._core.QgsSingleBandGrayRenderer object at 0x7f471c1da8a0>
Single band raster layers can be drawn either in gray colors (low values = black, high values = white) or with a pseudocolor algorithm that assigns colors to the values. Single band rasters with a palette can also be drawn using the palette. Multiband layers are typically drawn by mapping the bands to RGB colors. Another possibility is to use just one band for drawing.
Single Band Rasters¶
Let’s say we want a render single band raster layer with colors ranging from
green to yellow (corresponding to pixel values from 0 to 255).
In the first stage we will prepare a
QgsRasterShader object and configure
its shader function:
1 2 3 4 5 6 7
fcn = QgsColorRampShader() fcn.setColorRampType(QgsColorRampShader.Interpolated) lst = [ QgsColorRampShader.ColorRampItem(0, QColor(0,255,0)), QgsColorRampShader.ColorRampItem(255, QColor(255,255,0)) ] fcn.setColorRampItemList(lst) shader = QgsRasterShader() shader.setRasterShaderFunction(fcn)
The shader maps the colors as specified by its color map. The color map is provided as a list of pixel values with associated colors. There are three modes of interpolation:
Interpolated): the color is linearly interpolated from the color map entries above and below the pixel value
Discrete): the color is taken from the closest color map entry with equal or higher value
Exact): the color is not interpolated, only pixels with values equal to color map entries will be drawn
In the second step we will associate this shader with the raster layer:
renderer = QgsSingleBandPseudoColorRenderer(rlayer.dataProvider(), 1, shader) rlayer.setRenderer(renderer)
1 in the code above is the band number (raster bands are
indexed from one).
Finally we have to use the
to see the results:
Multi Band Rasters¶
By default, QGIS maps the first three bands to red, green and blue to
create a color image (this is the
MultiBandColor drawing style.
In some cases you might want to override these setting.
The following code interchanges red band (1) and green band (2):
rlayer_multi = QgsProject.instance().mapLayersByName('multiband') rlayer_multi.renderer().setGreenBand(1) rlayer_multi.renderer().setRedBand(2)
In case only one band is necessary for visualization of the raster, single band drawing can be chosen, either gray levels or pseudocolor.
We have to use
to update the map and see the result:
Raster values can be queried using the
sample method of
You have to specify a
and the band number of the raster layer you want to query. The method returns a
tuple with the value and
False depending on the results:
val, res = rlayer.dataProvider().sample(QgsPointXY(20.50, -34), 1)
ident = rlayer.dataProvider().identify(QgsPointXY(20.5, -34), QgsRaster.IdentifyFormatValue) if ident.isValid(): print(ident.results())
In this case, the
method returns a dictionary, with band indices as keys, and band values as
For instance, something like