Sometimes, you need to create a temporary data set for quick output or as an intermediate step in a more complex operation without the overhead of actually writing a file to disk. PyQGIS employs memory layers that allow you to create a complete vector data set, including the geometry, fields, and attributes, virtually. Once the memory layer is created, you can work with it in the same way you would work with a vector layer loaded from disk.
This recipe entirely runs inside the PyQGIS console, so no preparation or external resources are required.
We will create a Point vector layer, named Layer 1 with a few fields and then validate it:
- Start QGIS.
- From the Plugins menu, select Python Console.
- In the Python console, create a
QgsVectorLayer, including fields, and specify it as a memory data provider:vectorLyr = QgsVectorLayer('Point?crs=epsg:4326&field=city:string(25)&field=population:nt', 'Layer 1' , "memory") - Now, validate the layer and ensure that the console returns
True:vectorLyr.isValid()
The QgsVectorLayer requires three arguments. The last argument specifies the type, which in this case is memory. The second argument specifies the layer name. Normally, the first argument is the path to the file on disk, which is used to create the layer. In the case of the memory layer, the first argument becomes the construction string for the layer. The format uses query parameters that follow the convention key = value. We first specify the coordinate reference system and then specify the fields we want. In this case, we specify the first field, a string for city names, and then an integer field for population.
You can easily see how describing a layer's attribute table structure in a string can become unwieldy. You can also use a Python-ordered dictionary to build the string dynamically, as shown in the following steps.
- First, you need to import the
OrderedDictcontainer, which remembers the order in which keys are inserted:from collections import OrderedDict - Then, build an ordered dictionary that contains attribute names and types:
fields = OrderedDict([('city','str(25)'),('population','int')]) - Next, build a string by joining the output of a Python list comprehension that loops through the ordered dictionary:
path = '&'.join(['field={}:{}'.format(k,v) for k,v in fields.items()]) - Finally, use this string to define the layer:
vectorLyr = QgsVectorLayer('Point?crs=epsg:4326&' + path, 'Layer 1' , "memory")