Page 1 of 1

reading RATS output

Posted: Sat Jun 07, 2014 1:04 pm
by Nabtheberliner
Tom,
Actually i have the same pb with cov matrix of residuals from the same example from the TSAY book, i get the following output with RATS:

Code: Select all

Covariance\Correlation Matrix of Coefficients
         LDRUK        LDRCA        LDRUS
LDRUK 0.0000283715   0.09259709   0.23138784
LDRCA 0.0000026790 0.0000295021   0.43249284
LDRUS 0.0000073900 0.0000140854 0.0000359526

 2.83715e-005
 2.67895e-006  2.95021e-005
 7.39004e-006  1.40854e-005  3.59526e-005
The reading isn't clear, in the book they obtain directly

Code: Select all

    LDRUK        LDRCA        LDRUS
LDRUK 0.283715   
LDRCA 0.026790 0.295021   
LDRUS 0.073900 0.140854 0.359526
With the RATS output, how am i supposed to know that i have to skip the first four 0 after the dot?
Thanks again Tom.
Nabih

Re: reading RATS output

Posted: Sat Jun 07, 2014 1:46 pm
by TomDoan
If you multiply the data by 100 (which I would recommend anyway), the variances will be multiplied by 10000, so you will be spot on.