Hi,

I need to fit some data points to a function: y = A.(1 - exp(-B.x))
and get the best-fit values of A and B.

It seems excel comes with some push button fitting functions (linear,
polynomial) but not custom ones. Is this the case? If so, what do
people generally use to fit their data?

Thanks,
Ciao.

Put trial values for A and B in two cells.
Create a new column with an estimated value (based on the A and B cells)
for each observation in the data set.
In another cell, calculate:
- the sum of squared differences between actual and estimated
observations (for least squares estimation) or
- the sum of absolute differences between actual and estimated
observations (for least absolute deviation estimation) ...
Use Solver to minimize that sum by changing the values of A and B.

Jerry

trouble wrote:

> Hi,
>
> I need to fit some data points to a function: y = A.(1 - exp(-B.x))
> and get the best-fit values of A and B.
>
> It seems excel comes with some push button fitting functions (linear,
> polynomial) but not custom ones. Is this the case? If so, what do
> people generally use to fit their data?
>
> Thanks,
> Ciao.

"Jerry W. Lewis" <post_a_reply@no_e-mail.com> wrote in message news:<3F930458.705@no_e-mail.com>...
> Put trial values for A and B in two cells.
> Create a new column with an estimated value (based on the A and B cells)
> for each observation in the data set.
> In another cell, calculate:
> - the sum of squared differences between actual and estimated
> observations (for least squares estimation) or
> - the sum of absolute differences between actual and estimated
> observations (for least absolute deviation estimation) ...
> Use Solver to minimize that sum by changing the values of A and B.
>
> Jerry
>

Damn, is that the only way? How am I going to teach that to non-mathematicians???

Write a macro?

Jerry

trouble wrote:

> "Jerry W. Lewis" <post_a_reply@no_e-mail.com> wrote in message news:<3F930458.705@no_e-mail.com>...
>
>>Put trial values for A and B in two cells.
>>Create a new column with an estimated value (based on the A and B cells)
>>for each observation in the data set.
>>In another cell, calculate:
>>- the sum of squared differences between actual and estimated
>>observations (for least squares estimation) or
>>- the sum of absolute differences between actual and estimated
>>observations (for least absolute deviation estimation) ...
>>Use Solver to minimize that sum by changing the values of A and B.
>>
>>Jerry
>>
>>
>
> Damn, is that the only way? How am I going to teach that to non-mathematicians???