QGIS - Avoid duplicates on Value-Relation widget

Question

In QGIS, I'm working with two layers that are related to each other through the value-relation widget.

Layer 1: TEST

• fields: "id"; "LIB"; "odc_ID"; "Code"

Layer 2: ASSO_TEST

• fields: "id"; "LIB"; "CORRESP"

The value relation widget is on ASSO_TEST, field LIB

I'm attempting to generate a list containing one occurrence of each unique value in the LIB field, based on the corresponding values in the Code field within the ASSO_TEST layer.

---

I encounter an issue when there are multiple features with identical labels, leading to duplicate entries in the value-relation list.

In a previous project, I was able to filter this list using the following expression:

id = array_first(array_agg("id", group_by:= "Label"))

However, in this current project, I'm encountering difficulties and I'm unsure of what I might be doing wrong.

I have tried several approaches, like :

"Lib" = (array_distinct(array_agg("Lib", group_by:='odc_ID'))) 
"Lib" IN (array_agg("Lib", group_by:= 'odc_ID'))
"odc_ID" IN (array_distinct(array_agg("odc_ID", group_by:= "code")))
array_find("odc_ID",array_distinct(array_agg("odc_ID", group_by:= "code")))
"Lib" IN (array_to_string(array_distinct(array_agg("Lib", group_by:= "code")), ', '))
array_contains(array_distinct(array_agg("odc_ID", group_by:= "code")),"odc_ID")

I wouldn't have tried out so many variations of the code if I properly understood what I'm doing. But for my previous projects, I didn't have issues.

Here is a link to download a test project.

Test project

• she_weeds: You were overthinking it, if you want unique values of LIB then the expression is "id" = array_first(array_agg("id",group_by:="LIB")) - just change the group_by parameter. It's already answered here gis.stackexchange.com/a/431609/98784 and in other places. The filter can be very slow on large datasets. Also, not sure what you meant by filtering LIB based on Code field in ASSO_TEST?... there is no Code field in ASSO_TEST.

Answer 1

As explained by she_weeds, it was indeed very simple:

"id" = array_first(array_agg("id",group_by:="LIB"))