data(Block)
Block
block.names=colnames(Block)
block.names
## Not run:
# Block.m.1<-CircE.BFGS(R=Block,v.names=block.names,m=1,N=48,start.values="IFA")
# # You will get the following warnings:
# # WARNING!
# # INPUT COVARIANCE/CORRELATION MATRIX IS NOT POSITIVE DEFINITE.
# # STARTING VALUES CANNOT BE COMPUTED USING 'IFA': SET start.values='PFA'
# # Error in ifa(R, k) :
# # Make sure the listwise, not pairwise, missing data treatment has been selected
# # in computing the input matrix
# #
# # The error message reported by CircE highlights that the problem in fitting a CSPMF
# # to Block's data occurs at the early stage of starting values computation.
# # CircE offers a way to sidestep this problem without actually trying to discern its cause:
# # In fact, at this stage, the IFA cannot be carried out since the inverse of the
# # non-positive definite Block's matrix has negative values
# # on the diagonal (Joreskog, 1969, p. 54):
#
#
# diag(solve(Block))
#
# # To overcome this problem, CircE offers an alternative ways to obtain initial
# # parameters estimation, based on Principal Factor Analysis (PFA):
#
# Block.m.1<-CircE.BFGS(R=Block,v.names=block.names,m=1,N=48,start.values="PFA")
# ## End(Not run)
Run the code above in your browser using DataLab