How to estimate a VAR(k) model with diagonal autoregressive and covariance matrices using varm?

I used to use vgxvarx with 'CovarType','Diagonal' to estimate a VAR model with diagonal autoregressive and covariance matrices in R2017a.

But vgxvarx is no longer used in a newer version, instead varm and estimate are used.

Here is my code:

----------------------------------

irow = 213; % number of series

dt = logical(eye(irow));

VAR1diag = vgxset('ARsolve',repmat({dt},1,1),'asolve',true(irow,1));

VARdiag = vgxset(VAR1diag,'nAR',k,'ARsolve',repmat({dt},K,1));

[EstSpec,EstStdErrors,LLF,W] = vgxvarx(VARdiag,Yest,[],Ypre,'CovarType','Diagonal');

My revised code with varm follows:

-----------------------------------

irow = 213; % number of series

dnan=diag(nan(irow,1));

VARdiag = varm('AR',repmat({dnan},1,k));

[EstSpec,EstStdErrors,LLF,W] = estimate(VARdiag,Yest,'Y0',Ypre);

When I run the revised code, MATLAB said, requested 363378x363378(983.8GB) array exceeds maximum array size preference...

I think there is a similar property option like 'CovarType','Diagonal', but can't find it.

If you know how to fix this issue, please let me know. I appreciate your help.

Thanks.