Confirmatory factor analysis (CFA) of multivariate lipid distributions across time and race-ethnic groups: United States, 2003-2012

1. Run mplus-results.Rmd in R

   • Prior to running this program run export-mplus.R to get data in Mplus format using MplusAutomation R package.
     • Prior to running this program, get NHANES data in the read-data.R program, which pulls NHANES data off of the web.
   • Use templates with MplusAutomation package (createModels function) to make a batch of scripts to run each of the four models by year, race and sex groups.
     • Template for model 1
     • Template for model 2
     • Template for model 3
     • Template for model 4
   • Run models in Mplus using MplusAutomation in batch mode in virtuallab.unc.edu.
     • Log into virtual.lab.unc.edu
     • Open Mplus
     • Open Windows File Explorer
       • Type in %userprofile% to get to folder Mplus access on the remote site
         • Note: I had to contact support on 9/23/2016. A special version of R on virtuallab was set up for me to be able to call Mplus in virtuallab.
       • Open new file explorer and open local folder specified in Template for model 1, currently 'C:/temp/models2'.
       • Transfer .inp files from c:/temp/models2/ local folder to remote folder so can run MPlus program in virtuallab.unc.edu
     • Open R in virtuallab.unc.edu and run run-models.R. This will run all the Mplus programs one after another.
     • Got to remote folder and transfer the .out files to the local folder to save the Mplus output.
   • Take Mplus output and read in all output using extractModelParameters and extractModelSummaries functions in MplusAutomation in the mplus-results.Rmd program.
   • Make tables in R with the Mplus results now in R data frame formats.

2. Create poster and e-poster (really slides) file

   • aha-2017-cfa-slides.Rnw -- e-poster file
   • aha-2017-cfa.Rnw -- poster file