asked on

Interpolating unknows from polynomial curve

Hi,

I'm am trying to script an existing workflow and I am struggling with the last part.

The users have a standard curve of points as part of the data (x = percentage, y=value) and in an external software (Prism, GraphPad) utilise a 6th order polynomial non-linear regression to utilise that curve and then interpolate unknowns from that.

I have my x and y in 2 arrays ready to go.
In trying to recreate this I have been having a try with Math::Polynomial but the interpolate function of that seems to be using the curve data to predict the polynomial needed so I am a little confused.

Does anyone have a hint as to how best to approach this in perl by reading in curve data then reading out the unknowns?

Thanks

ozo

If you give Math::Polynomial->interpolate 7 data points, it can generate a 6th order polynomial that passes through those points, you can then evaluate that polynomial at other points.
If that's not what you want, perhaps you want to use Algorithm::CurveFit or PDL::Fit::Polynomial

Epipgx

ASKER

Thanks Ozo,

So there isn't a way to specify the the nth order? It is (as reading a bit more on the subject I think it should be) driven from the curve itself and produces the best order fit for your data?
If I have understood that correctly can you point me to an example where once the curve is fit using that value to read unknowns from it?

ASKER CERTIFIED SOLUTION

ozo

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

Epipgx

ASKER

Thanks Ozo that is fantastic, you have gone way beyond for me there and is much appreciated.

Epipgx

ASKER

Fantastically thorough answer, I was hoping for guidance but received a fully formed solution to my question, brilliant