conversion software: from GAUSS to RATS

[raob]

Dear All,

I shall be grateful for information on conversion software. I find it lot easier to work with RATS than with GAUSS. Thefore, I am looking for conversion software to convert SRCs for GAUSS into SRCs for RATS.

Bill

[TomDoan]

I've generally found that it's better to translate rather than transliterate Gauss or Matlab code. Particularly if you're doing something which allows for a variable number of inputs (number of lags in a VAR, number of explanatory variables), the coding in one of the matrix languages can be quite messy, while RATS can often handle this by inputting an equation and using the %EQN... functions to manipulate it.

(BTW, the biggest problem with a lot of Gauss/Matlab code is the lack of comments. I've often spent longer trying to figure out what this block of ten lines does than it takes to rewrite them once you have it figured out).

If you have something specific, I can take a look at it and maybe give you some pointers.

[raob]

Thank you Tom. Indeed I am grateful for your help with the Bai-Perron tests. While I made these tests work with both RATS (with your revised version of the SRC for RATS) and GAUSS, the latter's output gives results on several F tests. But then, in my sample, these F tests were always insignificant. So, I have to depend on the SBIC to determine the number of breaks, which RATS gives anyway.

I also agree that translation software creates (more) problems e.g., conversion of plain TeX documents into Word documents, although this is good for LaTeX documents. I will be in touch with you about some problems I am facing with another set of structural break tests. Perhaps a few other applied economists who might have used these structural break tests would share their difficulties. When I contacted someone who used these tests for more information, he/she was mum! Later I discovered that he/she has used GAUSS. My experience with GAUSS is only 3 weeks.