anybody knows what schema it is?

[#1]

Hii.. I am new to qv..

my problem is.. I make a link table on my qlikview project and the schema become like this.

link table is begin with %

anybody knows what schema it is?

star schema or snowflake schema?

I think this is a snowflake, but I still doubt about this...

thanks

[Clever_Anjos]

If you have two or more degrees of links between you tables, you have a snowflake

[#3]

so.. this is a snowflake, right? thankyou Clever Anjos

[maxgro]

to me it looks like a star

Dimension table

Star Schema Only have one dimension table for each dimension that groups related attributes. Dimension tables are not in the third normal form.

Snowflake Schema May have more than 1 dimension table for each dimension due to the further normalization of each dimension table.  Dimension tables are in the third normal form (3NF).

[phaneendra_kunc]

That is definitely a SNOW FLAKE....

Is there a reason to create that Link table and dimensions tables connected to second set of dimension tables?

I know you are making a Link table to get to your fact table....By looking at your model it is not scalable to Large data sets..

you might see performance issues when you deal with large data..

Try to avoid #of hops between your dimension tables to get to the fact....try to get more like a STAR SCHEMA...

Just a suggestion ...

[#6]

i'm confused.. T__T

[Anonymous · Accepted solution]

Not everything is easy to classify.

This schema looks exceptionally familiar.  It was invented by Phil Bishop who introduced me to QlikView.  I'd call it a quasi-star.  No, it is not an "official" classification.  Yes, Clever Anjos, from its shape it looks like snowflake...

What you see on the left are dimensions.  The tables %_dim_ are in fact the second-level links between dimensions and Master Link table, their role is to allow work with facts records that don't have values for some dimensions and don't turn these facts to 0s.  These auxiliary tables can be joined with their dimensions tables - the only difference that the dimension tables will be doubled in size.  But in that case it'll look more like a star.

The Fact_ tables to the right of the Master Link are the parts of a classic Fact table.  They could be safely concatenated in one large Fact table.  So, Master Link and Fact_ together are logically is the center of this "star".

The table on the right is another dimension table, for some reason implemented inconsistently, it looks out of place here.  It doesn't go to the Facts through the Master Link as other dimensions.  This doesn't affect the classification of this data model, but it affects the behavior of the application - if make selections in this dimension, all facts except Kuesioner will be 0.  (Unless you use set analysis to prevent this.)

Regards,

Michael

[#8]

Thank you  phani221984  

i just avoiding synthetic key and looping so I make the link table...

If i change my schema like this (still snowflake schema,right?)..

which one is better?

or you have any suggestion so I can move dim_pertanyaan to the left  (I mean connect dim_pertanyaan to master link table) without any synthetic or looping? so the schema can be a star schema

[maxgro]

many thanks Michael, your explanation is exceptional