The associative model show when you have more than one table.
Create a new table with, lets say, classes:
studentid classname start end
1 TIC 2012-01-01 2012-03-31
2 TIA 2012-01-01 2012-03-31
1 TID 2012-01-01 2012-03-31
Then studentid will associate with your first table. Compare with SQL JOIN (but be careful, it's NOT like SQL JOIN).
You don't have to create the link, QlikView does this whenever the column name match.
Ok thomas, how the data would be stored in associative database. where the columns would be and the data would be?
as we know that it will store the distinct values of every item in the table. then the data would be stored uniquely right?
then how will they relate the data with the columns and where the column names are stored.
I'm just a developer :-), not that familiar with the strict technical side of QlikView. I pretty much don't care how it's done, I just accept it and work accordingly.
As a developer the best thing for you to do is try. Load a bunch of data and play with it. Attend the, free, online classes and download, for free, QlikView.
If you on the other hand really need this indepth knowledge I must refer you to someone else. Probably QlikTech...
I've been a QlikView developer for six years. I have never needed to know the in depth structure of QlikView's associative data model. Your superiors don't need to know it either, even if they think they do.
It's like evaluating Excel for your company, and your superiors asking you, "How does Excel really store the data? What is its internal data structure?" The answer is that nobody cares, and nobody needs to care. All that matters is that it works. There are vastly more relevant questions to be asking when evaluating a software tool.
That said, your superiors probably want what they want, and some random guy on a forum saying "they don't need to know that" isn't likely to dissuade them. Fortunately, the basic ideas of how the data is stored internally are not that complicated. I think I posted this already in another thread of yours, but here's pretty much everything I know about the actual internal data storage.
Associative Database Management systems have been around since 1962..and predate SQO by a decade.. DR Codd quantified the levels of efficiency of Data Normalization with his 1971 paper. By defining levels 1 through 6, we have since then settled on 3rd normal form as no DB exists today except AtomicDB , which is an (N) normal form.. hence 4 dimensions more efficient.. and in DATA management higher efficiency = Speed .. This will explain it Associative DataBase Technology Review - YouTube