Displaying 6 results from an estimated 6 matches for "fafn".
Did you mean:
afn
2007 May 19
2
What's wrong with my code ?
I try to code the ULS factor analysis descrbied in
ftp://ftp.spss.com/pub/spss/statistics/spss/algorithms/ factor.pdf
# see PP5-6
factanal.fit.uls <- function(cmat, factors, start=NULL, lower = 0.005,
control = NULL, ...)
{
FAfn <- function(Psi, S, q)
{
Sstar <- S - diag(Psi)
E <- eigen(Sstar, symmetric = TRUE, only.values = TRUE)
e <- E$values[-(1:q)]
e <- sum(e^2/2)
e
}
FAgr <- function(Psi, S, q)
{
Sstar <- S - diag(Psi)
E <- eigen(Sstar, symmetric = TRUE)...
2006 Jul 19
2
get rid of error in Factor Analysis
B??dnie zakodowany tekst zosta? usuni?ty...
Plik: nie znany
Url: https://stat.ethz.ch/pipermail/r-help/attachments/20060719/3a883e4f/attachment.pl
2007 Jul 13
2
nearest correlation to polychoric
...2002 Computing the nearest correlation matrix - a problem from finance, IMA Journal of Numerical Analysis (2002), 22, 329-343.
which basically works but still leaves very small negative eigenvalues which causes factanal() to fail with
> factanal(cov=ncor$cor, factors=2)
Fehler in optim(start, FAfn, FAgr, method = "L-BFGS-B", lower = lower, :
L-BFGS-B ben?tigt endliche Werte von 'fn'
Zus?tzlich: Warning message:
NaNs wurden erzeugt in: log(x)
> version
_
platform i386-pc-mingw32
arch i386...
2005 Jun 26
0
Factor correlations in factanal
...unction(Psi, S, q)
{
sc <- diag(1/sqrt(Psi))
Sstar <- sc %*% S %*% sc
E <- eigen(Sstar, symmetric = TRUE)
L <- E$vectors[, 1:q, drop = FALSE]
load <- L %*% diag(sqrt(pmax(E$values[1:q] - 1, 0)), q)
diag(sqrt(Psi)) %*% load
}
FAfn <- function(Psi, S, q)
{
sc <- diag(1/sqrt(Psi))
Sstar <- sc %*% S %*% sc
E <- eigen(Sstar, symmetric = TRUE, only.values = TRUE)
e <- E$values[-(1:q)]
e <- sum(log(e) - e) - q + nrow(S)
## print(round(c(Psi, -e), 5)) # for traci...
2006 Jul 21
0
(no subject)
Hi,
I am having problems when trying to apply factor analysis in a covariance matrix,
> factanal(covmat=strip1cmcd, factors=5, control=list(lower=0.0264))
Error in optim(start, FAfn, FAgr, method = "L-BFGS-B", lower = lower, :
non-finite value supplied by optim
In addition: Warning message:
NaNs produced in: sqrt(diag(cv))
I have searched for possible solutions in messages posted previously. I think that:
Zeroing the workspace is not a requirement of t...
2008 Sep 09
1
Addendum to wishlist bug report #10931 (factanal) (PR#12754)
...unction(Psi, S, q)
{
sc <- diag(1/sqrt(Psi))
Sstar <- sc %*% S %*% sc
E <- eigen(Sstar, symmetric = TRUE)
L <- E$vectors[, 1:q, drop = FALSE]
load <- L %*% diag(sqrt(pmax(E$values[1:q] - 1, 0)), q)
diag(sqrt(Psi)) %*% load
}
FAfn <- function(Psi, S, q)
{
sc <- diag(1/sqrt(Psi))
Sstar <- sc %*% S %*% sc
E <- eigen(Sstar, symmetric = TRUE, only.values = TRUE)
e <- E$values[-(1:q)]
e <- sum(log(e) - e) - q + nrow(S)
## print(round(c(Psi, -e), 5)) # for traci...