Page 1 of 1

Estimation of Large-Scale BVARs

Posted: Fri May 29, 2009 4:30 am
by rangangupta
Hello,
I have been trying to estimate a large-scale BVAR comprising of 120 quarterly series. However, I keep on getting the following error message:
## I1. Expected Instruction - V65 Is Not Recognizable As One
>>>> v65 <<<<


Would really appreciate some help. I am attaching both the data and the code that I am using. In case other parts of the code require modification as well, please do make the appropriate changes as well.

Thanks
Rangan

Re: Estimation of Large-Scale BVARs

Posted: Fri May 29, 2009 2:25 pm
by TomDoan
Your input lines are way too long. Break them into more manageable pieces with something like:

Code: Select all

variables  v1	v2              v3	 v4	v5	v6	v7  $	v8	v9	v10	v11	v12	v13	v14 $	v15	v16	v17	v18	v19	v20	v21 $
                v22 ....
$ is the continuation character. If you leave it off and just break a line, RATS assumes that the next line is the start of a new instruction, hence your message. RATS will also break the line itself once it gets too long (500+ characters).

Note also that that's far too many variables to sensibly handle with the amount of data that you have.

Re: Estimation of Large-Scale BVARs

Posted: Sat May 30, 2009 1:17 am
by rangangupta
Thanks a lot, will give it a go. I will use the monthly version of this data set.

Regards
Rangan