Subscribe Bookmark RSS Feed

Issue with Analysis, Screening, Explore Missing Values

PeterL

New Contributor

Joined:

Oct 5, 2017

Hi,

 

I am trying to impute missing values in JMP Pro 13 through the new, Analysis, Screening, Explore Missing Values.

 

Its seems to work just fine, until I click "OK" to the warning that I should save my data to a new data table.

I can see the imputed values in the data table.

 

So, I click "OK" to the warning to get back to biz, and the imputed values disappear and are replaced with JMP's missing value code.  

 

When I repeat the process, the platform tells me that I don't have missing values, even though there are missing values in the data table.

 

Does anybody else have issues with it?  If so, what am I missing?

 

BTW, the old, Analysis, Multivariate Methods, Multivariate,  Red Diamond, Impute Missing Data works fine, but the new Screening platform seems a slicker.

 

Thanks!

 

 

 

2 REPLIES
ih

Community Trekker

Joined:

Sep 30, 2016

I often use that platform without issue so if you are having trouble you might need to open a support ticket. When I run the script below and press 'Multivariate Normal Imputation', then 'Yes, Shrinkage' and finally 'OK' for the data-in-memory-changed error, the value 49 persists in row 10.  Does that value disappear with this dataset and is this the workflow you refer to?

 

//Open sample data and set a value to missing
dt = Open("$SAMPLE_DATA/Fitness.jmp");
Column("Oxy")[10] = .;

mv = dt << Explore Missing Values(
	Y( :Age, :Weight, :Oxy, :Runtime, :RunPulse, :RstPulse, :MaxPulse ),
	Missing Value Report,
	SendToReport(
		Dispatch( {}, "Missing Columns", OutlineBox, {SetHorizontal( 1 )} )
	)
);

 

PeterL

New Contributor

Joined:

Oct 5, 2017

Thanks!
I found the issue. My variables have formulas. So Explore Missing Values
does not over write the cell values.
On the other hand, Multivariate Methods, Multivariate, Impute Missing
Values creates a new data table and leaves the old data table unchanged.

Makes sense.