What is “Associative”?
Qlikview is Associative in the way that the data is structured and able to be manipulated by the user. Unlike a hierarchical or query-based system, Qlikview stores ALL data and makes it available to the user for display. As the user selects a data point(or points), that selection is highlighted in green. Data associated with the selection becomes highlighted in white. Lastly, data that is not associated with the selection is highlighted in Gray.
As soon as the user wants to explore new data, they can simply click on any of the values. Association colors may change, but the data is still available for selection at any time, from any point.
In contrast, query-based Business Intelligence tools require a user to select a series of parameters and then “run” the report. If the user wants to make a change, they are forced to start over, as the query does not present data that is unrelated to the original selections.
Suppose you arrive at the grocery store and make your way to the bread section. Once you have the bread, you realize that you’ll need some butter. You don’t need to re-enter the store to find the butter. You would simply navigate the store from section to section. You would not return to the front door every time you had a new question, or discovered a new need.
Since its implementation in Qlikview 10, this feature has made it simpler to make selections based on associative relationships between data points. You may have noticed that if you select a list box and click the search icon or simply start typing, a search box appears. This search box allows us to narrow down results very quickly within the field displayed. The feature makes it very handy to make selections in long lists, without having to scroll through many data points.
But the search box is able to do so much more. This is because of the associative search feature built into the search. When the chevron icon is pressed, the search is expanded from the current field to all fields (by default) in the data model.
As soon as the button is pressed, a list is expanded showing the results searched in all fields. This search goes beyond the list boxes displayed on the screen, and can search through all fields in the dataset.
It is important to note that any selections made in this list will ONLY IMPACT THE CURRENT FIELD VIA ASSOCIATED RELATIONSHIPS.
In other words, selecting the city of “Jupiter”, for example, would not be the same as making the selection in a corresponding “City” list box. Instead, selecting this city will narrow my results in the “Month” listbox where I am currently searching, and will return only those months associated with sales made in the city of Jupiter.
This is what happens when I select Jupiter
I can see that November is the only month associated with sales in the city of Jupiter. However, only if I then press “ENTER” will this selection be made.
If I had instead clicked anywhere else on/off the screen, my search would be reset.
At this point, the selection “November” was made in the Months field, but the city field remains without a selection. So I would see sales for November across all cities.
This feature makes it simple to narrow down results for a multitude of different possible business questions, without cluttering the interface with dozens of list boxes that would not normally be used.