Creating a row vector of function handles

I have pre-defined function handles f_{i} using cell arrays for say i =1 to 8, so total 8 function handles. Now how do I create a new function handle, say g, which is a row vector consisting of these 8 function handles?

[Stephen23]

Function handles are the only fundamental data type that cannot be concatenated into arrays. They are scalar only:

https://www.mathworks.com/help/matlab/matlab_prog/fundamental-matlab-classes.html

Multiple function handles can be stored in container class arrays, e.g. structure arrays, cell arrays.

Answer 1

[Simar]

Hi Saurabh,

I understand that you want to create a new function handle, which should be a row vector consisting of pre-defined function handles.To define a new function handle ‘g’ as a cell array of function handles ‘f_{i}’, here is a simple way to do this:

% Assuming f_{i} are pre-defined 
f{1} = @(x) x^2;  % Example function handle 
f{2} = @(x) x^3;  % Example function handle 
f{3} = @(x) x^4;  % Example function handle 
f{4} = @(x) x^5;  % Example function handle 
f{5} = @(x) x^6;  % Example function handle 
f{6} = @(x) x^7;  % Example function handle 
f{7} = @(x) x^8;  % Example function handle 
f{8} = @(x) x^9;  % Example function handle 
% Create new function handle g 
g = @(x) cellfun(@(f) f(x), f, 'UniformOutput', false); 

Here, ‘g’ represents an anonymous function. This function takes one input ‘x’ and applies it to each function handle in the cell array ‘f’. The ‘cellfun’ function applies a function to each cell in a cell array. The function being applied is ‘@(f) f(x)’, which takes a function handle ‘f’ and applies it to x.

The 'UniformOutput', ‘false’ argument is necessary because the functions in ‘f’ might not all return outputs of the same type or size. If all functions in ‘f’ are guaranteed to return scalars, you can omit this argument to get a numeric array output instead of a cell array. You can call ‘g’ with a scalar input like this:

results = g(2);  % Apply all functions in f to the number 2 
results will be a 1-by-8 cell array where each element is the output of one of the functions in ‘f’ applied to 2. 
results =  
[4]    [8]    [16]   [32]    [64]    [128]    [256]    [512] 

To return a row vector instead of a cell array, modify the definition of ‘g’. In this case, g(2) will be a 1-by-8 numeric array.

g = @(x) cell2mat(cellfun(@(f) f(x), f, 'UniformOutput', false)); 
results = g(2) 
results =  
[4,   8 ,  16,  32,   64,   128,   256,  512] 

Kindly refer to the documentation links below for further information:

• Create Function Handle - https://in.mathworks.com/help/matlab/matlab_prog/creating-a-function-handle.html?searchHighlight=function%20handle&s_tid=srchtitle_support_results_1_function%20handle
• Apply Function to each cell in cell array (MATAB ‘cellfun’)- https://in.mathworks.com/help/matlab/ref/cellfun.html?s_tid=doc_ta
• Covert cell array to ordinary array of underlining data type (MATLAB ‘cell2mat’)- https://in.mathworks.com/help/matlab/ref/cell2mat.html?searchHighlight=cell2mat&s_tid=srchtitle_support_results_1_cell2mat

Hope it helps!

Best Regards

Simar