How to reorder output string in terms of values from a QTableWidget?

Question

I have a QTableWidget which contains the names of layers and values. The code below creates a range of values from a list, iterates through some groups whilst creating a dictionary containing the layer name and value from the QTableWidget.

The output is a simple SQL expression which contains the layer name and its value in order of when the layer appears in the table. I want to change this order so that the layer with the higher values are included first.

Here is the code I used:

def func(self):
    root = QgsProject.instance().layerTreeRoot()
    group_name = root.findGroup('Group')
    qTable = self.dockwidget.tableWidget
    delim = ' WHEN '
    # Number of layers in 5 groups
    layer_count = [15, 14, 1, 2, 1]                  

    top = list(np.cumsum(layer_count))
    bottom = list(np.subtract(np.cumsum(layer_count),layer_count))
    pairs = zip(bottom,top)
    ranges = iter([range(x[0], x[1]) for x in pairs])

    # Iterate through the 5 groups
    for group in group_name.children():
        value_dict = {}    
        for row in next(ranges):
            first_item = qTable.item(row, 0)
            first_name = str(first_item.text())
            second_item = qTable.item(row,1)
            value = str(second_item.text())
            value_dict[first_name] = value

        layers = [g.layerName() for g in group.children()]
        fscorestrs = {k:delim.join('"{layer}_Score" = {N} THEN "{layer}_Score"'.format(
                layer = l, N = value_dict[first_name]) for l in layers) for k in range(1)}
        formula = ("CASE WHEN " + fscorestrs[0] + " ELSE NULL END")

---

For example:

Here is a simple layout for the QTableWidget:

If I use the above code, the formula reads:

CASE 
WHEN "Layer_1_Score" = 0.1 THEN "Layer_1_Score" 
WHEN "Layer_2_Score" = 0.7 THEN "Layer_2_Score" 
WHEN "Layer_3_Score" = 0.5 THEN "Layer_3_Score" 
WHEN "Layer_4_Score" = 0.8 THEN "Layer_4_Score" 
WHEN "Layer_5_Score" = 0.6 THEN "Layer_5_Score" 
ELSE NULL 
END

What I am looking for is to instead produce an output where the layers are ordered in terms of the value, so the above should look like:

CASE 
WHEN "Layer_4_Score" = 0.8 THEN "Layer_4_Score" 
WHEN "Layer_2_Score" = 0.7 THEN "Layer_2_Score" 
WHEN "Layer_5_Score" = 0.6 THEN "Layer_5_Score" 
WHEN "Layer_3_Score" = 0.5 THEN "Layer_3_Score" 
WHEN "Layer_1_Score" = 0.1 THEN "Layer_1_Score" 
ELSE NULL 
END

How could I achieve this?

Answer 1

Firstly, import this library at the beginning of the code:

import operator

Then, once you have defined the layers list and assuming that:

delim = 'WHEN ' # the difference is only on the blank space before the 'W' letter

you may use this code:

# sort 'value_dict' by values
sorted_value_dict = sorted(value_dict.items(), key=operator.itemgetter(1), reverse = True)
fscorestrs = delim.join('"{layer}_Score" = {N} THEN "{layer}_Score"\n'.format(
                layer = sorted_value_dict[w][0], N = sorted_value_dict[w][1]) for w in range(0, len(sorted_value_dict)))
formula = ("CASE"+"\n"+"WHEN " + fscorestrs + "ELSE NULL" +"\n" + "END")

instead of what you have provided. I also slightly edited the final formatting of the expression, hoping you like it.

Using random values as a test, I get these prints:

print value_dict
{'layer_1': 0.3457999606457025, 'layer_3': 0.8843901948820355, 'layer_2': 0.6233214287240216, 'layer_5': 0.4854387498039854, 'layer_4': 0.07955152029877766}

print sorted_value_dict
[('layer_3', 0.8843901948820355), ('layer_2', 0.6233214287240216), ('layer_5', 0.4854387498039854), ('layer_1', 0.3457999606457025), ('layer_4', 0.07955152029877766)]

print formula
CASE
WHEN "layer_3_Score" = 0.884390194882 THEN "layer_3_Score"
WHEN "layer_2_Score" = 0.623321428724 THEN "layer_2_Score"
WHEN "layer_5_Score" = 0.485438749804 THEN "layer_5_Score"
WHEN "layer_1_Score" = 0.345799960646 THEN "layer_1_Score"
WHEN "layer_4_Score" = 0.0795515202988 THEN "layer_4_Score"
ELSE NULL
END