Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.



Re: IMSL error in CVF
Posted:
Sep 13, 2001 4:16 PM


In article <3BA0FDA6.DF79620C@263.net>, Dale <lidale@263.net> wrote:
>I met the following error message when using the routine DBVPFD if the >IMSL library to solve an >ODE system. I looked at the help file but can not find what it is. >Anyone has some idea? The NINTV is not an argument of DBVPFD.
>Thanks
>Dale >
>*** TERMINAL ERROR 1 from DPPDER. The number of intervals must be at >least >*** 1 while NINTV = 1 is given. > Here is a traceback of subprogram calls in reverse order: > Routine name Error type Error code >    > DPPDER 5 1 (Called >internally > DCSVAL 0 0 > DB2PFD 0 0 (Called >internally > DBVPFD 0 0 > USER 0 0 >
The traceback is telling you that DBVPFD called DB2PFD internally. This eventually led to a call to DCSVAL, perhaps from one of your callback routines. Are you using DCSVAL to evaluate the derivatives that are required by DBVPFD? DCSVAL then called DPPDER internally, which is where the error was detected. If you look up CSVAL/DCSVAL and PPDER/DPPDER in the chapter on Interpolation and Approximation, you will find that these routines compute cubic splines and their derivatives, and each of them has an argument called NINTV representing the number of polynomial pieces to be used in the spline approximation. The error message says there should be at least one interval.
It appears that DCSVAL is being called from your code. You should probably check to see what value is being passed for NINTV.
 Dave Seaman dseaman@purdue.edu Amnesty International calls for new trial for Mumia AbuJamal



