topic Standardized plot in graph builder with spec limits in Discussions

Standardized plot in graph builder with spec limits

Mauro_Gerber — Fri, 06 Mar 2020 09:57:44 GMT

When I drag and drop multiple coulumns with diffrent spec limits in the Y or X axis, JMP makes something like a standardized plot with point or boxplots.

It's a very nice feautre, but I could not find the documentation how the blue (spec) and gray (reference frame) lines are calculated.

Thanks!

Re: Standardized plot in graph builder with spec limits

txnelson — Fri, 06 Mar 2020 15:27:20 GMT

I have not been able to replicate your findings. What release are you using?

Re: Standardized plot in graph builder with spec limits

Mauro_Gerber — Mon, 09 Mar 2020 06:35:10 GMT

Hello Jim

Attached is a sample file with the script who produces the plot.

Thanks

Re: Standardized plot in graph builder with spec limits

txnelson — Mon, 09 Mar 2020 07:59:56 GMT

What appears to be happening, is that when you drag in as many columns to the X drop area, Graph Builder is taking on the format of a Parallel plot, but without the connecting lines, just the default points. So what you need to do is to look at the documentation for the definition of the graph.

Re: Standardized plot in graph builder with spec limits

Mauro_Gerber — Tue, 10 Mar 2020 08:59:19 GMT

THX, but I could't find something about this plot in the documentation.

Re: Standardized plot in graph builder with spec limits

txnelson — Tue, 10 Mar 2020 14:18:45 GMT

If you are referring to the horizontal blue lines, they are not calculated. They are taken directly from the USL, Target and LSL values in the column properties for the columns specified in the graph. The gray vertical lines are also not calculated but are just reference points for each column.

The documentation can be found in

Help=>JMP Documentation Library=>Essential Graphing=>Graph Builder=>Parallel

Re: Standardized plot in graph builder with spec limits

Mauro_Gerber — Mon, 16 Mar 2020 06:22:18 GMT

THX for the reply I read trough the doc and I played a bit with the spec and distribution. I think that the different scale of the spec limits is due to rounding of the individual axis.