I oversee several sections of WeBWorK for Applied Calculus and there are templates that were written about 3 years ago dealing with finding the antiderivatives of a binomial expression raised to an integer power. The templates were written with the answer evaluator function_cmp_up_to_constant. These templates seemed to be working fine using WeBWorK 2.1(?), but, recently, after upgrading to the current version, some of the instructors have informed me that some of these templates generate the error message "Can't solve for adaptive parameters". Doing a search through the forums, I see that the same message is generated when dealing with ODEs. What does the error mean in the context of finding antiderivatives? and is there a simple way of correcting the error?

Thank you.

Sherman Wong

### Antiderivatives and "Can't solve for adaptive parameters"

by Sherman Wong -
In reply to Sherman Wong
Tuesday, 20 April 2010, 9:12 AM

### Re: Antiderivatives and "Can't solve for adaptive parameters"

by Gavin LaRose -
Hi Sherman,

Davide is the best person to answer this question, but my experience has always been that I get the error when the system is trying to match functions on a poorly chosen domain. The result is that the functions are ill-behaved and it becomes difficult to find a domain on which the test points give a meaningful comparison.

The solution for the problems I've had that produced this error was to double check that the limits for the function comparison were chosen to lie in a domain where I was sure that the functions involved were well behaved (non-singular and well bounded).

Gavin

Davide is the best person to answer this question, but my experience has always been that I get the error when the system is trying to match functions on a poorly chosen domain. The result is that the functions are ill-behaved and it becomes difficult to find a domain on which the test points give a meaningful comparison.

The solution for the problems I've had that produced this error was to double check that the limits for the function comparison were chosen to lie in a domain where I was sure that the functions involved were well behaved (non-singular and well bounded).

Gavin

In reply to Gavin LaRose
Tuesday, 20 April 2010, 10:52 AM

### Re: Antiderivatives and "Can't solve for adaptive parameters"

by Sherman Wong -
Gavin:

Thank you for the suggestion.

Sherman

Thank you for the suggestion.

Sherman

In reply to Sherman Wong
Tuesday, 20 April 2010, 9:53 AM

### Re: Antiderivatives and "Can't solve for adaptive parameters"

by Davide Cervone -
There was another report of this recently that I fixed just a few days ago. Try updating your pg/lib/Value/Formula.pm file to the latest HEAD version (then restart the apache webserver) and see if that resolves it for you. The problem manifests itself with the FormulaUpToConstant object, but the error was in the underlying Formula object.

Davide

Davide

In reply to Davide Cervone
Tuesday, 20 April 2010, 10:51 AM

### Re: Antiderivatives and "Can't solve for adaptive parameters"

by Sherman Wong -
Davide:

Thank you very much for the suggestion. I'll forward it to the technician who oversees the server and the program.

Sherman

Thank you very much for the suggestion. I'll forward it to the technician who oversees the server and the program.

Sherman