Getting RAND() value but NOT function?!?

Hi everyone,

Say I am using a linear term such as " b * X "; where "b" is a
parameter while "X" is a variable...


Now, assume that "b" is RAND() value and assume I am using an excel
optimizer. When I run the optimizer, it complains that the problem
doesn't satisfy linearity conditions; because RAND() is a function so
it turns "b * X" to become non-linear!!!


I tried to go around this by creating another cell which copy and
value-paste the RAND() value into it. It solves the problem, but now I
have different problem: when RAND() value is copied and value-pased
and, by the time it is put in the new cell, the original cell which has

RAND() function changes its value.....and so on!!


How can I go around this correctly, so I keep the model linear and keep

RAND() values in both cells same?


Thanks,
Mike

Mike,

If you want to optimize a function, then you could put a random number in that cell as a seed, but
you need to use constants otherwise. It really doens't make sense to optimize a randomized
function, since the optimization would be (perhaps) different for every random number.

So, copy the cell with b, and pastespecial values, then run the optimizer.

HTH,
Bernie
MS Excel MVP


"Mike" <[email protected]> wrote in message
news:[email protected]...
> Hi everyone,
>
> Say I am using a linear term such as " b * X "; where "b" is a
> parameter while "X" is a variable...
>
>
> Now, assume that "b" is RAND() value and assume I am using an excel
> optimizer. When I run the optimizer, it complains that the problem
> doesn't satisfy linearity conditions; because RAND() is a function so
> it turns "b * X" to become non-linear!!!
>
>
> I tried to go around this by creating another cell which copy and
> value-paste the RAND() value into it. It solves the problem, but now I
> have different problem: when RAND() value is copied and value-pased
> and, by the time it is put in the new cell, the original cell which has
>
> RAND() function changes its value.....and so on!!
>
>
> How can I go around this correctly, so I keep the model linear and keep
>
> RAND() values in both cells same?
>
>
> Thanks,
> Mike
>

"Mike" <[email protected]> wrote in message
news:[email protected]...
> Hi everyone,
>
> Say I am using a linear term such as " b * X "; where "b" is a
> parameter while "X" is a variable...
>
>
> Now, assume that "b" is RAND() value and assume I am using an excel
> optimizer. When I run the optimizer, it complains that the problem
> doesn't satisfy linearity conditions; because RAND() is a function so
> it turns "b * X" to become non-linear!!!
>
>

You do not understnad what linear means. If b is indeed a parameter - fixed
value - then you are OK. If b is a random number then it is not a fixed
value and the equation is no longer linear.


--
Peter Aitken

Remove the crap from my email address before using.