Capability Analysis JMP Add-In

[#1]

Purpose

This add-in is designed to produce an attractive graphical rendition of a process capability study, and to make it easy to perform analyses across multiple columns of data.

How to Use

• Convenient navigation is delivered using a "wizard metaphor".
• Distribution fitting is automatically performed.  You can control the alpha level at which choices are made (including the ability to force  a normal fit).
• Buttons on the main window allow you to drill-down and view the "native" JMP output via the distribution platform.
• Specification limits are retrieved from column properties and can also be added / amended manually.
• Process capability statistics can be summarised across all columns in both tabular and graphical formats.
• Easily save / export images of your analysis.
• Interactive dashboard with colour-coded icons for each variable and sliders that allow you to control Cp and Cpk thresholds.  The dashboard is linked to the main window (click on an icon and the main window will update).
• Supports the use Z-Score as an alternative to the Percentile method when calculating capability indices for non-normal data (see 'non-normal distribution options' under the red triangle).
• The thumbnails can be arranged vertically or horizontally. The vertical arrangement is designed to sit next to the main window whilst the horizontal layout is intended to fill the full screen. Hover over a thumbnail to view Cp/Cpk indices. Click on a thumbnail to view in the main window.

System Configuration

Add-in developed and tested using the following system configuration:

JMP 17

JMP 16 

It has not been tested on versions prior to version 16.

[David_Burnham]

As with app-stores, ratings are a useful mechanism for users to voice their requirements and for developers hear what is and what is not useful.  Based on number of viewings this is probably my most popular addin - but is it actually used or useful - with zero ratings and comments I have no idea.

As the release of version 12 of JMP approaches I'll be making a decision as to whether or not I will discontinue supporting this add-in.  Currently I have no reason to continue with it, so please if you are using it and would like to have it maintained for future versions of JMP please let me know.

[kely_f_posada]

This is excelent! nice Add-in very useful since simplifies some steps when data is analyzed.

[David_Burnham]

Thanks kely.  It's been a time since I've updated it so if you find any problems or have ideas for improvements please let me know as I'll be looking at issuing an update over the next few months.

Dave

[mclayton200]

How about add-in for JMP standard edition?

[David_Burnham]

It ought to work with standard JMP

[Will_Penland]

Useful tool for process capability, thank you.   I'd like to see it report the Ppk indices as this has quickly become a more functional indices for larger sample sizes over time.

Kindly,

Will

[bharathu]

Does this work on JMP 13? 

[psunkari]

I have found it to be very useful tool, I think it would be advantage if option is provided to upload specficiations limits.

[Uriberm]

Did version 3 works for JMP16?

[DirtyDataCoLLC]

I really like this tool and have used it multiple times over the last couple of years. Thanks. 

[Neo]

@David_Burnham Does not seem to work in JMP 16.2. Any updates? Thanks

[David_Burnham]

Sorry, it's been a while since I've done updates.  I'll take a look at it and make sure it works for v16 & v17.

[Neo]

@David_Burnham Will eagerly wait for it.

[David_Burnham]

It's working on 16.1 and 17.0.  I'll track down version 16.2 and see if there is anything specific to that version.   OK, I can see what is going on now

[Neo]

@David_Burnham Thanks for looking into this. Will wait for your next update. 

[David_Burnham]

An update has been posted (version 3.1) but has to go through a review process so might be a few days before it is visible.

The purpose of the update is to provide compatibility with versions 16 and 17 of JMP.

I've also taken the opportunity to include support for the Z-Score method of calculating capability indices for non-normal data, which in my opinion is far superior to the default Percentile method.

Unfortunately the current distribution "fitters" are slower than their legacy counterparts (although they fit better!);  and the Anderson-Darling goodness of fit is painfully slow if you have large data.  In later versions I may resolve this by using a hybrid of p-value tests and model comparisons based on AICc.

Finally a comment about Cpk versus Ppk.  All indices are computed using "overall sigma".  I was going to add the option of either within or overall, but JMP only calculates overall sigma indices in the case of non-normal distributions.  Can anyone think of a reason for that?  I can't.  But it means if I want to provide support for both Cpk and Ppk formats I will need to implement the calculations myself - so that will be a later version.

[Neo]

@David_Burnham Not working in JMP 16.2. Halts at "Initializing cache..."

[David_Burnham]

@Neo , I've not been able to reproduce this error on JMP 16.2 (I've only tried on Windows 10).  Could it be data specific, is there an error message in the log?