turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- CPK Calculation - Ignoring a few values

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 14, 2017 3:03 AM
(756 views)

I am new to JMP and would need some help with CPK calculation.

At the completion of each experiment i have something around 600 values ranging from -5.0 to +5.0 (in an ideal scenario) and of these only values that are in the range of -3.0 to +3.0 are of interest to me. Is there an option to exclude the values that are outside the range -3.0 to +3.0?

Right now i am manually editing (as in deleting) the values in excel before importing them to JMP

Thanks in advance

3 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 14, 2017 5:48 AM
(740 views)

A direct but easier way is to select Rows > Data Filter. Select the data column of interest and click Add. You can now edit the range (shorten it) from the minimum and maximum of the values to the desired limits. Select the Include option near the top of the filter and the Distribution platform will only include data within the filtered range.

How is excluding data for computing the Cpk index justified? It is a measure of the process capability, so usually all of the process results are included?

Also, have you considered the time dependence of these values? Is the process in control while this data was collected?

Learn it once, use it forever!

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 15, 2017 12:29 AM
(720 views)

Sorry I should have been a bit more clearer.

Time dependence is not an issue and yes the process is in control. Each measurement value contains information about 3 coordinates (X,Y,Z) saved in 3 columns and if any of the coordinates fall outside the +3 to +3 range, the cooresponding X,Y or Z coordinates have to be discarded as well (Basically the whole row)

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 15, 2017 3:06 AM
(715 views)

This script illustrates Mark's suggestion above. Move the sliders and you should see the graphs update accordingly.

```
NamesDefaultToHere(1);
// Make some random data
dt = NewTable("Data",
NewColumn("X", Formula(RandomNormal(0,2.0))),
NewColumn("Y", Formula(RandomNormal(0,2.0))),
NewColumn("Z", Formula(RandomNormal(0,2.0))),
AddRows(100)
);
// Do capability with a local data filter
dt << Process Capability(
Process Variables( :X, :Y, :Z ),
Spec Limits(
X( LSL( -2 ), Target( 0 ), USL( 2 ) ),
Y( LSL( -3 ), Target( 0 ), USL( 3 ) ),
Z( LSL( -4 ), Target( 0 ), USL( 4 ) )
),
Goal Plot( 0 ),
Capability Index Plot( 1 ),
Local Data Filter(
Add Filter(
columns( :X, :Y, :Z ),
Where( :X >= -3 & :X <= 3 ),
Where( :Y >= -3 & :Y <= 3 ),
Where( :Z >= -3 & :Z <= 3 )
)
)
);
```