Pivot Grid Empty Cell Value/Filter Issue

[Mark]

I have a Pivot Grid, that has a data set that has two fields, In those fields, it's possible that records in that collection could have empty string value. And if I set one field to be my column group and the other to be my row group and one of those fields as my aggregate count field, everything looks fine, and as expected.   

However, as soon as I filter the ROW GROUP other than the EMPTY value, it drops the COLUMN groups "EMPTY" string column, as well, as if I filter the COLUMN GROUP on a non-empty value, it removes the ROW GROUP's empty string row.  I am not sure if this is by design or what but, this is causing an issue for our customers as they cannot figure out why their numbers in this pivot grid are not the same in other reports (non-pivot).  NOTE, if I filter in either the GROUP or COLUMN fields for the "EMPTY" string value, all the EMPTY string values still show in the COLUMN/ROW groups.  

As a solution to this problem, I made sure that these "EMPTY" string fields has at least one empty space in them... Attached is my sample project.  Once you run the form, just filter either the ROW or COLUMN group on anything other than the EMPTY string value.

 

Thanks

[Dinko | Tech Support Engineer]

Hello Mark,

Thank you for the provided image.

To avoid any misunderstanding in the setup, may I ask you to try attaching the sample project again as it seems it fails to attach? This way I can have a better look at how the RadPivotGrid is set and debug the reported behavior.

Regards,
Dinko | Tech Support Engineer
Progress Telerik

[Mark]

Hmm, I thought I did, let me re-add it.

 

[Mark]

Attached example

[Dinko | Tech Support Engineer]

Hello Mark,

Thank you for the provided project.

I was able to observe this behavior and I can confirm that this is unexpected. Therefore I have logged this on your behalf in our Feedback Portal where you can track its progress so that you will be notified by email for its progress. My investigations so far show that this is not related to the empty string but to the equal values. The filter algorithm does not respect correctly the scenario when you have equal values for both fields.

What I can suggest at this moment is to try to modify the source, if possible, to set a unique value for each field which will represent an empty string/value.

List<Class1> data = new List<Class1>
         {
            new Class1 { Field1 = "Empty Field1",  Field2 = "Empty Field2" },
            new Class1 { Field1 = "1", Field2 = "Empty Field2" },
            new Class1 { Field1 = "2", Field2 = "Empty Field2" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "A" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "A" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "A" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "A" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "A" },
            new Class1 { Field1 = "1", Field2 = "A" },
            new Class1 { Field1 = "2", Field2 = "A" },
            new Class1 { Field1 = "2", Field2 = "A" },
            new Class1 { Field1 = "3", Field2 = "A" },
            new Class1 { Field1 = "3", Field2 = "A" },
            new Class1 { Field1 = "3", Field2 = "A" },
            new Class1 { Field1 = "4", Field2 = "A" },
            new Class1 { Field1 = "4", Field2 = "A" },
            new Class1 { Field1 = "4", Field2 = "A" },
            new Class1 { Field1 = "5", Field2 = "A" },
            new Class1 { Field1 = "5", Field2 = "A" },
            new Class1 { Field1 = "5", Field2 = "A" },
            new Class1 { Field1 = "Empty Field1",  Field2 = "B" },
            new Class1 { Field1 = "1", Field2 = "B" },
            new Class1 { Field1 = "2", Field2 = "B" },
            new Class1 { Field1 = "2", Field2 = "B" },
            new Class1 { Field1 = "3", Field2 = "B" },
            new Class1 { Field1 = "3", Field2 = "B" },
            new Class1 { Field1 = "3", Field2 = "B" },
            new Class1 { Field1 = "4", Field2 = "B" },
            new Class1 { Field1 = "4", Field2 = "B" },
            new Class1 { Field1 = "4", Field2 = "B" },
            new Class1 { Field1 = "5", Field2 = "B" },
            new Class1 { Field1 = "5", Field2 = "B" },
            new Class1 { Field1 = "5", Field2 = "B" }
         };

I understand that this is will require extra work on your side, but still, you could try it and I hope it will work for you.

I have updated your Telerik Points for bringing this to our attention.

Regards,
Dinko | Tech Support Engineer
Progress Telerik

[Mark]

Yes, we have a work around for it, while not optimal, at least the grid is working as we would expect it to.

 

Thanks