Use colums in attribute table to set colors and line thickness directly automatic

Question

i have a column in my attribute table called 'Color' and 'Line' where I store rgb color codes and line thickness values (in map units) to give my layer elements a color and thickness. You can do it manually style => layerproperties.However I would like to use pyqgis to directly plot it properly.

Answer 1

Example for RGB column. Most of it from the answer to Applying catgorized symbol to each feature using PyQGIS, but with layer.getFeatures to read colors from attribute table. I have QGIS 3 so I had to change some of the syntax from pyqgis2 to 3, for example QgsSymbolV2 to QgsSymbol

layer = iface.activeLayer()
category_column = 'kommunnamn'
rgb_column = 'rgb'

color = {f[category_column]:f[rgb_column] for f in layer.getFeatures()} #Store all colors in a dictionary

fni = layer.fields().indexFromName(category_column)
unique_values = list(layer.dataProvider().uniqueValues(fni))
categories = []
for unique_value in unique_values:
    symbol = QgsSymbol.defaultSymbol(layer.geometryType())
    layer_style = {}
    layer_style['color'] = color[unique_value] #Read colors from the dictionary
    layer_style['outline'] = '#000000'
    symbol_layer = QgsSimpleFillSymbolLayer.create(layer_style)
    symbol.changeSymbolLayer(0, symbol_layer)
    category = QgsRendererCategory(unique_value, symbol, str(unique_value))
    categories.append(category)

renderer = QgsCategorizedSymbolRenderer(category_column, categories)
layer.setRenderer(renderer)
layer.triggerRepaint()

Answer 2

I've set a minimum project available at https://labs.webgeodatavore.com/partage/sample-datadefined-colors.zip. Unzip it. Open the qgs file with QGIS and use the following in the Python QGIS console

layer = iface.activeLayer()

# For color defined in attribute column
layer.renderer().symbol().symbolLayer(0).dataDefinedProperties().property(QgsSymbolLayer.PropertyFillColor).setExpressionString("\"rgb\" || ',255'")

layer.renderer().symbol().symbolLayer(0).dataDefinedProperties().property(QgsSymbolLayer.PropertyFillColor).setActive(True)

# For thickness in attribute column
layer.renderer().symbol().symbolLayer(0).dataDefinedProperties().property(QgsSymbolLayer.PropertyStrokeWidth).setExpressionString("\"thickness\"")

layer.renderer().symbol().symbolLayer(0).dataDefinedProperties().property(QgsSymbolLayer.PropertyStrokeWidth).setActive(True)

# Refresh rendering of the layer
layer.triggerRepaint()

You will just need to use the same recipe by looping on your various layers.

Unrelated to usage with PyQGIS but to choose the approach that "fit", the main drawback using data defined properties if that you don't get a "right" legend but you will be able to manage as many colors as you want. The other approach mentioned with QgsCategorizedSymbolRenderer is better for legend as long as you don't have hundred or thousand of categories (because predictable).