how to find the best curve fit for a set of data?

i have set of data points and want to know the best curve fit equation for this data . can any one explain me

[Image Analyst]

Do you want to include the flat parts in the fit?

Do you want a piecewise fit, like one for the flat parts and one for the curved parts?

Is there any particular model you'd like (exponential, polynomial, etc.)?

[data, strings, raw] = xlsread('matlab_data.xls')
x1 = data(:, 1);
y1 = data(:, 2);
x2 = data(:, 4);
y2 = data(:, 5);

subplot(2, 1, 1);
plot(x1, y1, 'b-', 'LineWidth', 2);
grid on;

subplot(2, 1, 2);
plot(x2, y2, 'b-', 'LineWidth', 2);
grid on;

[Deva Narayanan]

yes i need piecewise polynomials for both curved and flat model

Answer 1

[KSSV]

[num,txt,raw] = xlsread('Matlab_data.xls') ;
x1 = num(:,1) ; y1 = num(:,2) ;

p = polyfit(x1,y1,3) ;

x1i = x1 ;
yi = polyval(p,x1i);
y1i = polyval(p,x1);

plot(x1,y1,'r')
hold on
plot(x1i,y1i,'b')