Missing variables when running PCA

#1
Hi,
Kindly assist how to input for missing data when running PCA. Am trying to use Principal components analysis (PCA) to assign weights for each of my variables and Factor Analysis to generate the factor loadings for each of the variables. But when I run PCA, there are so many missing observations. Any help on how I can operate PCA successfully.
 

Mean Joe

TS Contributor
#2
If you run PCA on variables with missing data, beware that the components could be picking up a signal of missingness, in addition to the underlying factors that you are after.

Sorry that's all I've got to say. I tend to drop variables that have too many missing values, when doing PCA.