I am not a professional and sorry if I bring any incorrect concept. When you say impute, I guess you want to replace the missing part by its conditional expection.
This is not safe if you do non-linear opearations later. An simple example is the expection of quadratic form, E(X'AX) != E(X)'AE(X). In other words, for example, if you want to estimate the covariance, better avoiding using the imputed values directly since they will not give good results, but if you want to estimate the mean, then it is maybe OK (maybe, I do not remember my results). Please correct if I am wrong. I am also interested in those packages. But if you are stuck with finding a proper package, you may write your own. The most clear explaniation I have read is wiki. In order to use EM, you need to specify the distribution of the data, not sure if it is proper, depending what is your next step. And also, Bayesian methods and packages I think exist in R to impute the data. Best wishes, Jie On Mon, Jul 23, 2012 at 11:39 AM, David L Carlson <dcarl...@tamu.edu> wrote: > This is chopped off. What happens next is important. EM can be used to > compute covariance matrices and conducts the statistical analysis without > imputing any missing values or it can be used to impute missing values to > create one or multiple data sets. > > You might find Missing Data by Paul D. Allison to be useful. > > ---------------------------------------------- > David L Carlson > Associate Professor of Anthropology > Texas A&M University > College Station, TX 77843-4352 > > > > -----Original Message----- > > From: r-help-boun...@r-project.org [mailto:r-help-bounces@r- > > project.org] On Behalf Of ya > > Sent: Sunday, July 22, 2012 5:11 AM > > To: r-help > > Subject: Re: [R] EM for missing data > > > > hi Greg, David, and Tal, > > > > Thank you very much for the information. > > > > I found this in SPSS 17.0 missing value manual: > > > > EM Method > > > > This method assumes a distribution for the partially missing data and > > bases inferences > > on the likelihood under that distribution. Each iteration consists of > > an E step and an > > M step. The E step finds the conditional expectation of the b > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide > http://www.R-project.org/posting-guide.html<http://www.r-project.org/posting-guide.html> > and provide commented, minimal, self-contained, reproducible code. > [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
