Error: $ operator is invalid for atomic vectors

[MicheleCai]

Hello！
This error occurred when I modeled the MSGARCH model for the logarithmic return on Brent oil futures from January 4, 2011 to December 31, 2020. I demean this logarithmic return using AR(5). This problem does not occur when I use the sequential return of Brent oil futures over the same time interval to model. How do I solve this problem?

[KhalilBelghouat]

I'm finding the same problem! Did you find a work around?

[mark0100]

The same here.
The problem does not seem to be related to the format of the provided data to FitML() but rather to the data content.

I have lots of datasets to reproduce this problem with if someone is interested.

It looks like the problem is in line 131 of ../Package/R/FitML.R:

129 optimizer <- ctr$OptimFUN(vPw, f_nll, spec, data_, ctr$do.plm)
130
131 llk <- -optimizer$value

I suspect that the optimizer in some cases (depending on the actual data?) is returning an atomic vector and hence the error message "Error in optimizer$value : $ operator is invalid for atomic vectors".