Displaying fit function on the plot

Hello,

I have a fit function which is displayed below. There is a plot with this fitted function. Are there anyway that I can display the "f(x) = -0,02462x^2 - 8.336x -747.7" on the plot?

If someone know how to do it, pls help me.

Answer 1

[Steven Lord]

We can use some of the methods of the object returned by fit to get the pieces we need to display it.

x = (1:10).';
y = x.^2 + 2*x - 3 + randn(size(x));
P = fit(x, y, 'poly2')

P = 
     Linear model Poly2:
     P(x) = p1*x^2 + p2*x + p3
     Coefficients (with 95% confidence bounds):
       p1 =      0.9567  (0.8564, 1.057)
       p2 =        2.46  (1.328, 3.592)
       p3 =      -3.834  (-6.544, -1.124)

The formula method gives us the expression for the fit with the coefficient names.

F = formula(P)

F = 'p1*x^2 + p2*x + p3'

The coeffnames method gives us the coefficient names and the coeffvalues method the coefficient values.

N = coeffnames(P);
V = coeffvalues(P);

Now we can use string operations to replace the coefficient names with the coefficient values (converted to strings.)

formulaWithValues = replace(F, string(N).', string(V))

formulaWithValues = '0.95673*x^2 + 2.4599*x + -3.834'

That + followed by a - for the constant term is a little awkward looking. Let's fix that.

formulaWithValues = replace(formulaWithValues, "+ -", "- ")

formulaWithValues = '0.95673*x^2 + 2.4599*x - 3.834'

Now we could use formulaWithValues to add a legend entry for the fitted curve (or with the text function to put it in the axes itself.)

plot(x, y, 'o', DisplayName = 'raw data')
hold on
xx = 1:0.25:10;
plot(xx, P(xx), DisplayName = formulaWithValues)
legend show

This gets a little more complicated if your fit has problem-dependent parameters, but it wouldn't be that difficult to use probnames and probvalues in addition to coeffnames and coeffvalues.