HI All!
I am in the need of finding a graph that can visualize the "correlation strength" / (Dependency) between categories WITHIN 1 dimension and by using 1 measurement as strength value.
Example / Solution / Idea:
Could be a type of Network (Bubble) chart with (Strength/dependency) lines going between the bubbles making it look like a network.
Bubbles is the 1 Dimension (Products dimension "Article/item ID") and Strength lines is the 1 Measure (value) (Invoiced value). However, I guess there is needed 1 "relationship" dimension that ties the value and bubbles together. This would be the OrderNumber as the logic would be: If the products is in the basket/Order then tie those products (bubbles) together with invoice value.
Second option is a dependency wheel chart (But not liking this as much).
For The two charts just mentioned I have only found that they need 2 dimensions. one "from" dimension and one "to" dimension not being able to measure correlation of categories "within" only 1 dimension.
My e.g. case is regarding a product dimension and revenue measurement (could also for example be quantity or other measures).
The question is: of all (say) 100 000 orders/shopping baskets. How often do the same products appear together ?
E.g:
basket 1 = Apple of 10 eur and banana 5 eur and a radio 100 eur.
looking of all 100 000 baskets it should be seen that whenever apple is in the basket it is strong correlation to that banana is in the basket too but muck weaker (more seldom) that the radio is also in the basket.
Would be awesome to have like a big network chart with various strong "flows" / "Connection" between products. THis would show the "main" products that has many strong conenctions and islands of small products.
Anybody that can help with tips/advice on what chart to use, any extensions out there (hopefully free but ok if not)?
All advice are great!
.png "Luminary Alumni"){kind=icon}
