Does anyone know where I can find the code to create the correlation graph which was on the store in R and which is no longer there? In Vega or Vega lite for Deneb? Thank you so much.
@JimmyCruyp, thanks for the data that you provided. So, creating the viz in Deneb shouldn't actually be hard. What I'm having trouble with is calculating the values (correlation coefficient?) in the cells that are encoded as text and circles. I was able to calculate the correlation coefficient for the entire dataset. Admittedly I'm not great with statistics (something that I need to work on), so any insight into the formula for calculating the value for an individual cell would be greatly appreciated. Everything I'm finding is just instructions on how to create the chart in r.
@giammariam Hi, thanks a lot for your support. I really dont know. Have you tried to ask "Chat GPT"? Maybe he could support? At least for the understanding on the formula for calculation?
@JimmyCruyp, I finally found some time to get back at this. I actually found a really great guide by Parker Stevens on how to create a correlation plot with native visuals only. Screenshot from guide:
I then took his concept and incorporated DAX generated SVGs to include the circles and the values.
Here's the screenshot after my updates:
Let me know if this is enough to get you going or if you have any thoughts.
In the example, all queries are reference queries (besides _measures and Cars), with the base query being Cars
In your source data, if you have the variables in key-value pairs already, you wouldn't need the Cars query; you'd be starting with a query that looks like Cars Pivot.
If you have your variables as columns (like in the initial Cars query), then you'd flatten the values into key-value pairs, as was done in the Cars Pivot
Attributes, Attribues1, and Attributes2 are all single-column queries that just contain a unique list of the variables. One is for the x-axis of the matrix, and the other is for the y-axis.
The Attributes query is the bridge between Attributes1 and Attributes2, and is used to populate the slicer
Note - the cardinality and single cross filter direction are very important here
DAX
You'll want to modify the following measures to use your values (fortunately these measures are 99%" the same with the exception of how the return value is calculated):
n
X
Y
X^2
Y^2
XY
The only DAX variables that you'll want to make sure to change in the CC SVG measure are _w and _h to match the Image size formatting properties (see below screenshot)
Could you please just explain me the meaning of the different measures? If we take my example with currency, normallly you have only one measure right? What are all these measures for? Could you explain for each? THanks a lot, Jim
