What does fzero take as input for the function?

If i have a function that starts as follows:

function [y,index] = findsmode(x)

and if i then use

 sigma=fzero(@findmode,500);

then what does the function findmode take as input x?

Also, if i put

 disp(*)

within the function file with * being some variable calculated in the function, why does the value show up in my command window multiple times, while it is not being iterated.

I hope someone can explain this to me

Answer 1

[Steven Lord]

then what does the function findmode take as input x?

fzero will call your findmode function repeatedly with various values as it tries to search for a root of findmode.

You can kind of think of this like someone walking along a path. They look left one pace to determine if that location on the path is higher or lower than where they are, then look right one pace to determine the same thing, decide which way to step based on whether they're higher or lower than their destination, and repeat this process.

why does the value show up in my command window multiple times, while it is not being iterated.

You may not have a loop inside your function but fzero will call your function repeatedly as part of its search.

[matlabber]

Thanks!

Clear explanation. Do you perhaps know how I can visualize the various values it tries?

[Steven Lord]

Use optimset to set the 'Display' and/or 'OutputFcn' options.

fzero(@sin, 1, optimset('Display', 'iter'))

 
Search for an interval around 1 containing a sign change:
 Func-count    a          f(a)             b          f(b)        Procedure
    1               1      0.841471             1      0.841471   initial interval
    3        0.971716      0.825854       1.02828      0.856414   search
    5            0.96      0.819192          1.04      0.862404   search
    7        0.943431      0.809577       1.05657      0.870673   search
    9            0.92      0.795602          1.08      0.881958   search
   11        0.886863      0.775093       1.11314      0.897089   search
   13            0.84      0.744643          1.16      0.916803   search
   15        0.773726      0.698805       1.22627      0.941237   search
   17            0.68      0.628793          1.32      0.968715   search
   19        0.547452      0.520513       1.45255      0.993017   search
   21            0.36      0.352274          1.64      0.997606   search
   23       0.0949033     0.0947609        1.9051       0.94464   search
   24           -0.28     -0.276356        1.9051       0.94464   search
 
Search for a zero in the interval [-0.28, 1.9051]:
 Func-count    x          f(x)             Procedure
   24           -0.28     -0.276356        initial
   25        0.214567      0.212924        interpolation
   26     -0.00065816   -0.00065816        interpolation
   27     5.06172e-06   5.06172e-06        interpolation
   28    -3.62624e-13  -3.62624e-13        interpolation
   29     1.54849e-24   1.54849e-24        interpolation
   30     1.54849e-24   1.54849e-24        interpolation
 
Zero found in the interval [-0.28, 1.9051]

ans = 1.5485e-24

As far as I'm aware there isn't a predefined OutputFcn that plots x and the value of the function at the x value for each iteration but it is possible to write one. See this documentation page for more information.

[matlabber]

thanks a lot!