Subscribe Bookmark RSS Feed

run correlation in JMP and ignore missing values

odoyle81

Community Trekker

Joined:

Oct 26, 2011

How do I run a correlation in JMP and ignore missing values?  In SAS, I can use "nomiss"

Thanks in advance!

1 ACCEPTED SOLUTION

Accepted Solutions
Solution

in the multivariate platform, under the list of columns there is a drop down menu enabling you to choose Row-Wise.

Row-wise estimation does not use observations containing missing cells. The Row-wise option is useful if you want to check compatibility with earlier JMP versions (before JMP 8), if you want to explore the relationship of missing versus nonmissing cells within the dataset, or if you want to exclude any observations that have missing data. (Row-wise estimation was the only estimation method available before JMP 8.)

Pairwise performs correlations for all rows for each pair of columns having nonmissing values.

(the last section was copied from the help, you can get there by selecting the "?" tool and clicking on the output of the multivariate platform)

2 REPLIES
Solution

in the multivariate platform, under the list of columns there is a drop down menu enabling you to choose Row-Wise.

Row-wise estimation does not use observations containing missing cells. The Row-wise option is useful if you want to check compatibility with earlier JMP versions (before JMP 8), if you want to explore the relationship of missing versus nonmissing cells within the dataset, or if you want to exclude any observations that have missing data. (Row-wise estimation was the only estimation method available before JMP 8.)

Pairwise performs correlations for all rows for each pair of columns having nonmissing values.

(the last section was copied from the help, you can get there by selecting the "?" tool and clicking on the output of the multivariate platform)

odoyle81

Community Trekker

Joined:

Oct 26, 2011

perfect, thanks!