Correlate Y by X with Multiple Items

[#1]

Hi,

I have a table which consist of few parametrics item that I want to check for correlation with respect to few X items at one time. Can I know what is the best way to do it as per attached file. I would like to have param1 vs param2 and param3 at the same time for correlation check and compare against 2 different Eval, T_Type and No_ID. I think in Graph Builder, I can do that but how about in Analyze section. 

Thanks.

[#2]

Have you looked at 'Analyze > Multivariate Methods > Multivariate'? The script below runs this platform, using 'T_TYPE' (defined with Modeling Type Nominal) as a 'By' variable. Yyou can have more than one such variable.

Multivariate(Y( :PARAM1, :PARAM2, :PARAM3 ), By( :T_TYPE ));

[#3]

Hi Ian,
Thanks. Yes, I have tried that. Is it possible to have the By in this case the Eval,No_ID, T_Type appearing together(similar to MS Excel pivot chart.

Thanks.

[#4]

If I assign all three of those variable to the 'By' role, there is one point in each scatterplot matrix (but maybe this is just sample data).

Anyway, given your comment about piovot chart, I'm now not sure if you are looking for a table. or a graph, or both. There are lots of good reasons why the last choice would be the best.

So, can you describe further or give an example of what you want?

[#5]

Hi Ian,

I'm looking into something like pareto chart whereby my Y-axis will gives me param1/param2/param3 while my X-axis will show me the Eval, No_ID and T-Type. Please refer to example as per attached.

Thanks.

[#6 · Accepted solution]

After importing your data into JMP change the modeling type of T-Type and No_ID to Ordinal or Nominal and then drag all three of your columns (Eval, No_ID and T-Type) to the X axis in Graph Builder. Then drag your three parameters to the Y axis. You can shift-click on the Line element to add the line.

[#7]

Thanks Jeff and Ian. I guess this is what I'm looking for :)