# [Limdep Nlogit List] RE: Double Bounded (Probit)

William Greene wgreene at stern.nyu.edu
Sat Nov 17 00:32:25 EST 2007

```Ms. Abdulla.  The logminus that it refers to means that you are
trying to take the log of a nonpositive number in the expression.
This can happen if the data are such that the probabilities get
too close to 1.0 or 0.0, or in your case, too close to each other.
One other possibility is that there are missing values in your
data that are causing strange results in the probabilities.  You
might try using CREATE to compute the observations in the function
at the starting values, and see if the problem arises.  (MAXIMIZE
sums these values to obtain the function.)
/Bill Greene

----- Original Message -----
From: "Sabah Abdullah" <sabouha2 at gmail.com>
To: limdep at limdep.itls.usyd.edu.au
Sent: Thursday, November 15, 2007 9:42:29 PM (GMT-0500) Auto-Detected
Subject: [Limdep Nlogit List] RE: Double Bounded (Probit)

Dear LIMDEP users:

I am running a maximum likelihood estimate (MLE) using a double
bounded dichtomous choice (DB DC) format where there are four
combinations of choices (YY, NN, YN, NY). I ran the following commands
(see below) and a return error message (see below). What does these
ERROR messages mean?

-Thank you for your time and consideration.

Regards, Ms. Abdullah

sample; all\$
--> Maximize; labels=alpha,beta; start=0.25,0.25;
fcn=YY*log(1-lgp(alpha+beta*BID_HIGH))
+YN*log(lgp(alpha+beta*BID_HIGH)-lgp(alpha+beta*BID_INTL))
+NY*log(lgp(alpha+beta*BID_INTL)-lgp(alpha+beta*BID_LOW))
+NN*log(lgp(alpha+beta*BID_LOW))\$
Error:   590: Obs.=     1 Cannot compute function: Logminus
Error:   137: Iterations: function not computable at crnt. trial estimate
Error:   802: Cannot compute function at start values. Exit status=4.
Function=  .00000000000D+00, at entry,  .00000000000D+00 at exit
Error:   590: Obs.=     1 Cannot compute function: Logminus
Error:   137: Iterations: function not computable at crnt. trial estimate
Error:   143: Models - estimated variance matrix of estimates is singular
Error:   447: Current estimated covariance matrix for slopes is singular.
--> wald; labels=alpha,beta; start=b(1),b(2); var=varb; fn1=-alpha/beta\$
Error:    67: Matrix element in an expression has an invalid subscript.
Error:    32: Model command.  Start vals. unreadable or wrong # given.
Error:   208: MINIMIZE;...\$  Start values not given or contain bad values

***********************************************
Sabah Abdullah
PhD Student (Economics)/Part-time Tutor/Resident Tutor
Dept. of Economics & Int'l Development
BATH
BA2 7AY
UNITED KINGDOM
e-mail:	s.abdullah at bath.ac.uk
WWW:	http://www.bath.ac.uk/~sa245
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au

--
Professor William Greene
Department of Economics
Stern School of Business
New York University
44 West 4th St., Rm. 7-78
New York, NY   10012
http://www.stern.nyu.edu/~wgreene
212.998.0876

```