I have a shapefile with a column that have project references, I want to create a different shapefile for every project reference there is.

Note that the project references are not unique in the sense that there are multiple lines with the same reference.

Input example:

ID ref clmnC
1 001 abcd
2 002 addd
3 001 dsdd
4 003 sdsd
5 003 rzer
6 002 qdqd
7 001 ssqz

Output wanted (3 tables in this example):

  ref clmnC          ref clmnC           ref clmnC
  001 abcd           002 addd            003 sdsd
  001 dsdd           002 qdqd            003 rzer
  001 ssqz

I tried listing unique values but that only return a table with only unique project references values, any solution to solve this problem using the graphical modeler in QGIS 2.18?


2 Answers 2


I found the solution, using the 'Split Vector Layer' tool, I was able to split my datasets based on the attribute I wanted.



As an alternative using PyQGIS where Graphical Modeler is not required.

Let's assume there is a point layer 'points' with its corresponding attribute table accordingly, see image below.


Task: Split layer into several layers by "value" attribute and delete "id" field for each of it.

Proceed with Plugins > Python Console > Show Editor > New Editor and paste the script below

# imports
import os, sys

class SplitLayerToMany:
    # initiating layer's and field's names provided from user
    def __init__(self, layer_name, field_name):
        self.layer_name = layer_name
        self.field_name = field_name
    # setting up the input layer and defining required parameters
    def split_layer(self):
        # loading layer
        layer = QgsProject.instance().mapLayersByName(self.layer_name)[0]
        layer_fields = layer.fields()
        idx_of_that_field = layer_fields.indexFromName(self.field_name)
        if not layer.isValid():
            raise ValueError("Layer failed to load!")
        # getting a set with unique key attributes
        list_attributes = []
        for feat in layer.getFeatures():
        list_attributes = set(list_attributes)
        # checking a number of features in the input layer
        if layer.featureCount() < 2:
            raise ValueError("Layer contains only one feature. No reason to double save it")
            # getting a working dir, where the input layer is stored
            path_to_file = layer.dataProvider().dataSourceUri()
            working_dir = os.path.split(path_to_file)[0]
            # looping over attributes from the set, selecting features and saving them as a new single layer
            for i in list_attributes:
                selection = layer.selectByExpression('{0}=\'{1}\''.format(self.field_name, i))
                writer = QgsVectorFileWriter.writeAsVectorFormat(layer, working_dir + "/{}.shp".format(i), "UTF-8", layer.crs(), driverName = "ESRI Shapefile", onlySelected = True)
my_class = SplitLayerToMany('points', 'value')

Press Run script run script and get the output




Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.