The RATS Software Forum

Dear Tom Doan

I use two files(var_1_1.rpf and garch11_1.rpf, both are from Course Materials: Vector Autoregression, 2nd Edition and ARCH, GARCH, and Stochastic Volatility, 2nd Edition) to learn SVAR-MGARCH model.
The files have a same code
@VARLagSelect

Whenever I just run both files, I face the same situation
First, the window of file finding popup menu is showed
Then, then the error massage is showed
“## CP18. VARLAGSELECT is not the Name of a PROCEDURE. (Did you forget to SOURCE?)
>>>>@VARLagSelect(<<<<”

Except the code(@VARLagSelect), every code is worked in the files.

What’s wrong with my work?

With Catalina, the procedures are now put outside the applications folder, into (by default), the /Users/Shared/RATS/Procedures directory. Your preferences probably still are pointing to the old directory. You'll have to update it.

https://estima.com/ratshelp/preferences.html

Hi Tom,

I'm new to RATS. Just purchased my permanent license yesterday. I'm trying to use the Pedroni Panel Structural VAR ( Pedroni, Peter (2013) "Structural Panel VARs") in estimating a panel SVAR with 6 variables. i got the code from Prof. Pedroni but later also found it on Estima. However it seems @VARLagSelect code is not working. I follow your recommendation here but couldn't get it working. I understand the procedure has changed a little in the version 10. Can you please provide any other directions on how to get it working.
Thank you

Are you talking about the same issue with the "@VARLagSelect is not the name of a procedure" message?

Hi Tom,

The issue is with the @VARLagSelect showing as an error when you try running the code. However i was able to figure out that by using specified lags.
However there are other issues with some part of the Code for Pedroni (2013) Panel structural VAR not running smoothly on RATS version 10.0.
Is it possible to See what is wrong with the code.

I have attached my Code (which i'm still working on ) and the data. Will be great to get some assistance. Relatively new to coding and RATS but will do my best whiles waiting for help.

Thanks

First of all, I'm not sure what you're doing with the HP filters. FILTER(TYPE=HP) returns the HP trend itself. If you want the detrended data (I suspect you do), you need TYPE=HPDETREND. Also, that doesn't do a separate trend calculation for each cross section element, and I assume you want that. However, note that the overall analysis is already removing common time effects, so it's not clear that you want to add that separate detrending.

Thanks Tom. You are right. I was trying to detrend the series and use only the cyclical component. I will work on it again.
I'm grateful. Any chance you take a look at my code?

What Pedroni paper is that from?

It's From Pedroni, Peter (2013) "Structural Panel VARs", Econometrics, 1 (2),180-206.

Thanks for Your Help Tom. I got help from Prof. Pedroni to get it working.

Hello Francis!
Can you attend the final code from Pro. Pedroni?Thank you in advance.
I want to use the same method in my research.

Here you go man.

Hello Francis.
I am very happy for your sharing the code! When I run your code,besides the graph,I get the results:
"processing panel member 1
processing panel member 2
processing panel member 3
processing panel member 4
processing panel member 5
processing panel member 6
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs
## IO48. File with 577 columns might not be readable by other programs"

After I study your code,i find the code "open copy ind-VDs-oftotal-to-idiosyncratic-shocks.xls". However, i can't understand the sentence and where it comes from.
My problem is how to get ind-VDs-oftotal-to-idiosyncratic-shocks.xls.
Thank you in advance!

The program is outputting the impulse responses to XLS files. The old XLS spec allows for only 256 columns, and with the size of the model, it needs 577. It's harmless unless you want to look at those spreadsheet files. If you need those, you would need to switch from using XLS to XLSX.

Great input by Tom. Also, you should know the directory of your RATS. The files will be stored in there.

hello Francis
I understand what you said,and find the files.
thank you very much!