Least square fitting of data

Hi,

I have data(xdata,ydata) and want to fit a model with the data. The model is a generic one y=k/x, k is constant according to the shape of the data points. My questions are: 1. How can I know the precise form of the model? 2. After using lsqcurvefit, I get the constant. How can I get the equation itself to predict new y from new x.

Here is my data: xdata:[2;10;20;30;40;50;60;70;80;90;99.598] ydata: [98913.08;19922.26;10046.06;6751.64;5107.10;4116.99;3457.18;2988.11;2637.81;2358.34;2147.66]

Thanks in advance

Answer 1

[Star Strider]

Do this:

xdata = [2;10;20;30;40;50;60;70;80;90;99.598];
ydata = [98913.08;19922.26;10046.06;6751.64;5107.10;4116.99;3457.18;2988.11;2637.81;2358.34;2147.66];

eqn = @(k,x) k./x;                                      % Define Objective Function
B = lsqcurvefit(eqn, 1, xdata, ydata);                  % Estimate Parameter

xv = linspace(min(xdata), max(xdata));                  % Create New ‘x’ Vector
yv = eqn(B,xv);                                         % Use ‘B’ And ‘xv’ In ‘eqn’ TO Calculate ‘yv’

figure
plot(xdata, ydata, 'pg')
hold on
plot(xv, yv, '-r')
hold off
grid

[Abhay Athaley]

Thank you so much for the answer.

[Star Strider]

As always, my pleasure.