"The covariance of the parameters failed to invert"

GAUSS result:
return code = 2
maximum iterations exceeded

Log-likelihood -1.46942E+006
Number of cases 2058

Covariance of the parameters computed by the following method:
ML covariance matrix

The covariance of the parameters failed to invert

Parameters Estimates Gradient
params[1,1] 4.9834 -2412.3665
params[2,1] 0.0211-421828.8940
params[3,1] 0.24092071287.1206
params[4,1] 1.302625480245.6333

Number of iterations 100001
Minutes to convergence 16.96468

600.35536 303382.87 28607.935 352349.92
303382.86 1.5252354e+008 -1407620.5 -17336958.
-20026.871 -3501919.6 25863893. 2.1265951e+008
22858.627 4956390.8 -5551049.8 -83082630.


does this indicates the possible wrong with the specification of the mm.gradient??

1 Answer


Yes, most likely.   To see what is happening, set

struct cmlmtControl c0;

c0.gradCheck = 1;

CMLMT will then check the analytical gradient against the numerical gradient.  This will help you to find out if it's computed properly, and if not, which one of the calculations has the problem.

You must login to post answers.

Have a Specific Question?

Get a real answer from a real person

Need Support?

Get help from our friendly experts.

Try GAUSS for 14 days for FREE

See what GAUSS can do for your data

© Aptech Systems, Inc. All rights reserved.

Privacy Policy