# [LON-CAPA-dev] Installing Modules for LON-CAPA

**Gerd Kortemeyer
**
lon-capa-dev@mail.lon-capa.org

*Mon, 26 Sep 2005 16:10:47 -0400*

Hi Donnell,
On Sep 26, 2005, at 3:33 PM, GARRETT, DONNELL MAZON wrote:
>* Hi
*>*
*>* That was an example... maybe one that is too simple to show what I'm
*>* trying to do.
*>*
*>* Let's say we want to solve for x and y, and if we randomly generate A,
*>* B, C, D, E, and F, then
*>*
*>* xA + yB = C
*>* xD + yE = F
*>*
*>* how can we find x and y using an algorithm that we don't have to
*>* cut and
*>* paste into each and every problem.
*
Yep, you can use the .library mechanism (as Guy explained) to write
routines that you can then <import> into any problem.
But, what I was trying to suggest is building your problem the other
way around, i.e., randomly determine the solution first.
$x=&random(-4,4,3);
$y=&random(-4,7,3);
Then generate your coefficients:
$a=&random(4,8,1);
$b=&random(4,8,1);
$d=&random(5,9,1);
Now, when determining your $e, you just need to make sure that the
second equation is not a multiple of the first:
$e=$d/$a*$b+&random(4,8,1);
Then calculate your $c and $f:
$c=$a*$x+$b*$y;
$f=$d*$x+$e*$y;
To your students, you just give $a, $b, $c, $d, $e, and $f, of
course. Solving an equation system is computing-intensive,
constructing one is not - and there is no reason why LON-CAPA needs
to do the same work that the students do.
But maybe the example is really too simple.
- Gerd.