Help Fit line / curve to data

Hello, I am trying to find the equation of a best fit line for this data that I have below. I tried using polynomial fit with degree 2 but the results are unsatisfactory. I was wondering if anyone can offer any suggestions. Based on the scatter plot below, id like the results to be a curve that is steep at a steep slope at the beginning and gentle later.

[KSSV]

Have you checked lsline?

[RuiQi]

Hi thanks ! But id like to extract the complete equation of the curve as well.

[KSSV]

h = isline ;

h will be having the xdata and ydata.It fit's a straight line.

Answer 1

[Walter Roberson]

I suggest trying to fit it to y = A * exp(-B*x) + C

If you assume that C is 0 then that would be a fit of y = A * exp(-B*x) which would be

log(y) = log(A) - B * x

which would be a linear fit to find B and log(A).

[RuiQi]

Cool ! I would need to exponentiate log(A) so that I can get back A right ? And as for B I don't need to do anything to it

Answer 2

[KSSV]

doc lscurvefit.

[Walter Roberson]

lscurvefit needs to be told what model to fit against.