search

amices / mice

Multivariate Imputation by Chained Equations
https://amices.org/mice/
GNU General Public License v2.0
428 stars 107 forks source link

the plot of standard deviation (SD) of variables in the iteration cruve after multiple imputation are blank/empty #547

Closed bengi000 closed 1 year ago

bengi000 commented 1 year ago

Hello!

I was running mice package to mulitply imputed a dataset with missing values in variables BP_HIGH and BP_LOW. But after I drew the imputed object RR_mice, the sd (Standard deviation) of them in iteration curve were empty as below: image

I had changed the predictors many times or even imputed with no other predictors, but the results were same. and i check the sd of each dataset was below :

sd(result1$BP_HIGH) [1] 21.07227 sd(result2$BP_HIGH) [1] 21.07519 sd(result3$BP_HIGH) [1] 21.07592 sd(result4$BP_HIGH) [1] 21.07555 sd(result5$BP_HIGH) [1] 21.07336

and I noticed that the number of missing in these two variables were both 1, is it countable for the issue, or what else could be?

hanneoberman commented 1 year ago

Hi! This is expected behavior if there is only one imputed value for these variables. The plot shows the average and sd of the imputed values only. You could verify by looking at the $chainMeans and $chainVariance components of the mids object.

hanneoberman commented 1 year ago

Closing the issue because of expected behaviour in mice. For more information see https://stefvanbuuren.name/fimd/sec-algoptions