Goal Seek

[zdas04]

I just went through the goal seek exercise from hell trying to do a GENFIT exercise for six variables trying to match a data set stacked into 16 vectors (i.e., for each vector I would vary the guesses until one of the variables converged to a target number, capture the other 5 and move on the the next vector; at the end of the 16 vectors I would put the solved variable into the function and start over with the next one with one fewer variables). I guess I spent 30 hours tweaking guesses and trying again.

Now that it is over (the resulting equation matches the data set within 5%), I was wondering if there was any way that I could have automated it within native MathCad to do a goal seak like Excel does.

David

 

[PNachtwey]

I would use the minerr function and have it done in 15 minutes after I figured out how to read your data.
With only 16 vectors the actual computation time would be less than second. Using Scilab the results would be the same. I would use the optim function.