The RATS Software Forum

There is a an only graph there, it is the 5th one in my email. all of them are there
Rachid

It was labeled graph.rgf, so I assumed it was a graph.

Because of the recursive way that the data are generated, the two step process doesn't work unless you run it through the evaluation of the L frml. If you look at the creation of E:

set e = x-a10-a11*x{1}-a12*temp1{1}

what is temp1? At this point, all zeros. It isn't created until later so e is generated wrong, which cascades through the entire calculation. In short, the ugly graphs are the "correct" ones given your model. So you have to figure out why your MAXIMIZE estimates don't appear to be correct.

Dear Tom,

Thanks for your insightful and continuous support.
Rachid