Forum archive 2000-2006

Thomas Hagedorn - Another Problem Error

Thomas Hagedorn - Another Problem Error

by Arnold Pizer -
Number of replies: 0
inactiveTopicAnother Problem Error topic started 11/25/2003; 12:15:38 AM
last post 12/5/2003; 11:56:51 AM
userThomas Hagedorn - Another Problem Error  blueArrow
11/25/2003; 12:15:38 AM (reads: 938, responses: 6)

In Problem setIntegrals19Area/ns6_1_1.pg,   I believe there is an error.  The problem
wants one to calculate the area between two functions.   From the formula for the answer,
it appears as if the problem assumes that one function is always larger than the other. 

No problem, but in the specific example it gives my student,   it wants the area
from    x=-7 and x=4 between the graphs  of   f(x)=.1 x^2 + 2  and   g(x)=x (which
do intersect inside the region).   Looking
at the Problem Editor, I see how f(x) and g(x) are generated, but not how
the starting/ending points for x are generated (I'm expecting to see -7 and 4 as choices
for $b and $c, but don't see them.)   So I'm at a loss to correct the problem myself.

Thanks,
Tom

<| Post or View Comments |>


userArnold K. Pizer - Re: Another Problem Error  blueArrow
11/25/2003; 8:27:06 AM (reads: 1104, responses: 0)

Hi Tom,

According to the CVS,  this problem was updated about 19 months ago.  I'm not sure if this update fixed the bug you report.  Are you using the updated verion? In the new version line 97 reads

Find the area enclosed between  ( f(x) = $d x^2 ? {$a} ) and  ( g(x) = x ) from (x= $c  )  to (x= $b). $BR

(the from c to b is new).  This new version should have been contained in the 2002 and 2003 distributions of the problem library.  If you don't have it you are using very old problems (a remark I include for the benefit of other readers).

If you are using the current version, please send us the seed the causes the error you report and we'll try to fix this.

 

Arnie

<| Post or View Comments |>


userThomas Hagedorn - Re: Another Problem Error  blueArrow
12/1/2003; 6:38:41 AM (reads: 1075, responses: 0)
Arnie,

It looks like we have the updated problem (we only installed Webwork last July so that
problem should have had the corrected version) but I think I was unclear in my first post.   

The problem isn't with the statement of the problem.  The problem is that Webwork calculates the wrong answer for the specific example I mentioned above.   Most of the cases work just fine.  

In our problem, the answer given in the Webwork code is

$answer = ($d /3)*($b**3 -$c**3) + .5*($c**2 -$b**2) + $a*($b -$c);

but that isn't the correct answer when the functions intersect inside the given interval.  

-Tom

<| Post or View Comments |>


userArnold K. Pizer - Re: Another Problem Error  blueArrow
12/1/2003; 8:51:38 AM (reads: 1060, responses: 0)

Hi Tom,

If you can send us the seed used in the offending example, it will probably save us a lot of time hunting for the bug.  On the prof page, go to "Modify Problem Set For Student" where you will see the seed listed.  One way to quickly fix this type of bug for a single student is to use that page to change the seed but probably it's best to just give the student full credit [by putting a 1 in both the Attempted and Frac(tion) Corr(ect) boxes]. 

I understood your first post and quoted line 97 since that was one of several lines which were changed in the updated problem.  It's likely thoses changes were made to correct the bug you report but, if so, they were an unsuccessful attempt at a bug fix. 

Arnie

 

<| Post or View Comments |>


userThomas Hagedorn - Re: Another Problem Error  blueArrow
12/1/2003; 2:06:58 PM (reads: 1068, responses: 1)

I'll see if I can recreate the problem.   Once I discovered the error I deleted the problem from the problem set for the class, so I've "lost" the initial seeds.

Thanks,
Tom

<| Post or View Comments |>


userCoreen Mett - Re: Another Problem Error  blueArrow
12/5/2003; 9:08:55 AM (reads: 1208, responses: 0)

In looking at the latest version of this problem, I don't know a seed that will generate the difficulty that Tom describes.  But it is possible that the graphs intersect, and the answer presumes that they do not.  

The gist of the problem is to find the area between the functions f1(x) = d*x^2+a and f2(x) = x on the interval [b,c], where it is assumed that the graphs of f1 and f2 do NOT intersect.  The coefficient generator requires a in [2,8], b in [2,8], c in [-8,-2], and d in [.1,.95].  However, if a,b,d are ALL small, the graphs of the two functions WILL intersect on the interval [b,c].  This can easily be fixed by shifting f1(x) up one more unit.  Just change line 22 to "$a = random(3,10,1);" thereby assuring that the f1 > f2 on the entire domain.

Coreen

<| Post or View Comments |>


userArnold K. Pizer - Re: Another Problem Error  blueArrow
12/5/2003; 11:56:51 AM (reads: 1029, responses: 0)

Thanks for looking at this Coreen.  I have updated the problem in the CVS using your bug fix.

Arnie

<| Post or View Comments |>