Dear Alain, You may speed up the analysis by using the sample covariance matrix based on a listwise deletion: cov.cfa <- cov(your.raw.data, use="complete.obs")
Since you have 36671 cases, the results should be similar to those based on the raw data unless you have lots of missing data and/or the data are missing at random. By the way, if your questions are directly related to SEM, you may get more responses from SEMNET (http://alabamamaps.ua.edu/archives/semnet.html). Hope it helps. Regards, Mike -- --------------------------------------------------------------------- Mike W.L. Cheung Phone: (65) 6516-3702 Department of Psychology Fax: (65) 6773-1843 National University of Singapore http://courses.nus.edu.sg/course/psycwlm/internet/ --------------------------------------------------------------------- On Thu, Jun 2, 2011 at 1:21 AM, yrosseel <yross...@gmail.com> wrote: > On 06/01/2011 05:39 PM, D. Alain wrote: > >> Thank you Yves and Mike, >> >> your comments make sence, however they do not resolved my problem: The p >> < N is the result of my poor attempt to give a reproducible example. My >> "real" dataframe has a dim of 36671 cases an 41 variables. >> Following your advice, Yves, I passed my model without lavaanification >> (just "cfa(cfa.model, data=df.cfa,missing="ml")"), but now R is working >> for hours without printing any results... >> > > To monitor progress, you can use the verbose=TRUE argument. If you are not > fitting your final model, you may want to use se="none" (no standard errors) > and test="none" (no test statistic) to speed things up. > > I've ran analyses that took multiple days, using both lavaan and commercial > software. It is annoying (the wait), but not unusual. The time to fit is a > function of the number of missing patterns you have in your data, and the > number of variables. > > Yves. > [[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.
