sample weights

Umm, I'm not following the previous explanation. The definition variable "data.weight" is a 1 x 1 matrix, and "firstModel.objective" is a n x 1 matrix when vector=TRUE. Another way to specify this is to put the weights in a n x 1 MxMatrix, and then multiply the weights by the likelihood vector. I think there's a right-parenthesis missing the closing of the model "firstModel".

Hi,
Following this thread and a previous one [1], I've been trying to implement differential weighting of subjects in a model by building another model along side which multiplies the vector of fits in the first by the row weight in the data and optimizing that sum.

I thought it was working, but I am getting very different results from a colleague who is using MPlus's weight feature. Just deleting the low-weighted rows gives results more similar to his than to mine...
So,

wMZ = data.frame(weight = w[twinData$ZYG == "MZFF"]) # get weights for MZs
wDZ = data.frame(weight = w[twinData$ZYG == "DZFF"]) # and for DZs
...
# An MZ model with vector = T
mxModel(name = "MZ", mxData(mzData, type = "raw"), mxFIMLObjective("top.mzCov", "top.expMean", vector = bVector))

# model "MZw", which simply weights the MZ model
mxModel("MZw",
mxData(wMZ, "raw"),
mxAlgebra(-2 * sum(data.weight %x% log(MZ.objective) ), name = "mzWeightedCov"),
mxAlgebraObjective("mzWeightedCov")
# ...
# In the container model, target an algebra which sums the weighted objectives
# of the MZw and DZw, which in turn target weighted functions of the individual row
# likelihoods of the underlying MZ and DZ ACE models
mxAlgebra(MZw.objective + DZw.objective, name = "twin"), mxAlgebraObjective("twin")


What (if anything) is wrong with this approach? (or my implementation)

[1] http://openmx.psyc.virginia.edu/thread/445

With this example there aren't a lot of arrays available to be non-conformable. It's either -2, MZw.data.weight, or MZ.objective. First, I'd make sure the objective (fit function) for the MZ model has vector=TRUE. Second, is MZw.data.weight a single column vector? The * operator is for elementwise multiplication, and it will not recycle either argument if they are different lengths like R. Use %*% for matrix multiplication. Third, you could put MZw.data.weight into its own mxAlgebra and have a look at it that way.

So, building a simple play model. After encountering some no-context errors :-(
This is just estimating the covariance of X and Y, which is around .5

require(OpenMx)
require(MASS)
#Simulate Data
set.seed(200)
rs = .5
nSubs = 1000
selVars <- c('X','Y')
nVar = length(selVars)
xy <- mvrnorm (nSubs, c(0,0), matrix(c(1,rs,rs,1),2,2))
testData <- data.frame(xy)
names(testData) <- selVars
cov(testData)

# Now: Switch to vector
m1 <- mxModel(m1, mxFitFunctionML(vector = T), name = "vector")
m1 <- mxRun(m1)
umxSummary(m1, showEstimates = "std" )
# what we get
# name Estimate Std.Error
# 1 vector.expCov[1,1] 0.4705 9.0e+07
# 2 vector.expCov[1,2] 0.6148 1.4e+08
# 3 vector.expCov[2,2] 1.0314 2.1e+08
# what it should be...
cov(testData)
# XX 0.9945328
# XY 0.4818317
# YY 1.0102951


So when vector is on, something needs to change in the model to keep it driving toward good estimates?