Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

Dynamic classification

Hi everbody,

I want to classify following table but this table must be classify in a dynamic way i hope anyone knows how to do that:

Client;Client_Dist;K1_Dist;K2_Dist;(Sum(Data))

A;B;10;20;(30)

A;C;40;20;(60)

A;D;10;10;(20)

Sum(Data) is a defined forumla.

The client should classify by Sum(Data). Sum(Data) is the whole Data of my crosstable which i have define in QlikView. There are up to three key figures which you can choose.

For example:

Selection: Client (A); Client_Dist(B,C)!!!

(Client 'B') K1_Dist {10} and K2_Dist {20} -> Sum(Data) {30}

or

(Client 'C') K1_Dist {40} and K2_Dist {20} -> Sum(Data) {60}

                  Total Sum(Data) -> 90

Regarding to the selection you get different total Sum(Data) values. My goal is to classify the Clients into four cluster.

<=25 % of  Total Sum(Data); >25 and <= 50% of Total Sum(Data); > 50% and <= 75% of Total Sum(Data); > 75% of Total Sum(Data).

Result for our example is:

<= 25% (22,5); >25 and <= 50% (45); > 50% and <= 75% (67,5)

Client B would be in cluster 2 (>25 and <= 50%) -> 30

Client C would be in cluster 3(> 50% and <= 75%) -> 60

So is it possible to build such a dynamic classification in QlikView where you can define %values of the total sum fpr different cluster?

Thank,

Thomas

2 Replies
Anonymous
Not applicable
Author


Please post an example qvw , it will be helpful..

Not applicable
Author

Okay, on the right side you can see the "Cluster" diagram, in this diagram is a dimension which do the classification. But now the formula is not complete and got a mistake.

In the test table you can see

Sum(Daten) = 85;

Sum(Daten)25% = 21,25,

Sum(Daten)50% = 42,5

and so on ...

Now i need this values for my classification!

Example:

Kunde_Dist B got Sum(Daten)  = 25 and must be into Sum(Daten)50% cluster because 25 > 21,25 and <=  ...

I hope now you can understand what i mean. Maybe there is a solution regarding to the variables?!

Thank,

Thomas