Vector Autoregressive (VAR) - Inconsistency With Documentation

Dear All,

According to varm documentation,

"The innovations covariance matrix Covariance cannot contain a mix of NaN values and real numbers; you must fully specify the covariance or it must be completely unknown (NaN(NumSeries))."

The given example in this doc is:

"Example: 'Covariance',eye(2)"

I have a problem in specifying Covariance matrix just as the example:

y1 = randn(100, 1);
y2 = 2*y1 + randn(100, 1);
Y = [y1, y2];

ar_1 = nan(2);
mdl = varm('Lags', 1, 'AR', {ar_1}, 'Covariance', eye(2));

est_mdl = estimate(mdl, Y);

The raised error is:

Error using varm/estimate (line 385)

Covariance matrix of innovations must contain no restrictions.

I'm using MATLAB R2019a

[Viktoriya Glushko]

Hello, I have run into a similar problem and was wondering if you managed to solve yours? If yes, could you share what the solution was please? Many thanks. Viktoriya

[Dana]

I just ran OP's code snippet and didn't get any errors. Can you explain exactly what you're doing and where the error is showing up?

[ali shirali]

Hi,

I gave up on this issue at that time.

But after upgrading to MATLAB R2019b it seems to be solved. Now I'm working on Mac OS, MATLAB R2019b. The error was raised on Windows, MATLAB R2019a. I'm not sure is this a bug fixes or no.