How do you call data from a structure array in a function?

Hey all,

I'm new to MATLAB (and programming) and I have a question about how to call data stored in a structure array in a function. For a little bit of background, I have a bunch of related data stored in a scalar structure array I've named 'Data' with fields named f1, f2, f3 and so on that all contain 2x2 matrices.

To plot data from any of these fields seems simple enough, it's just (for example):

plot(Data.f1(:,1), Data.f1(:,2))

What I want to do now is create a function where the input(s) of the function dictate which of these fields are plotted. Thus, if the input variable of the function is x then the corresponding matrix to be plotted should be something like

plot(Data.f_x(:,1), Data.f_x(:,2))

However, when I try this is it obvious that MATLAB has no idea that x is supposed to be a variable. And the text in the function parenthesis isn't really a string (I guess) because I have tried using num2str() and sprintf() functions to try and get around this like

plot(Data.f_num2str(x)(:,1), Data.f_num2str(x)(:,2))

This just looks plain wrong to me, but I don't know enough to know why. Any help would be greatly appreciated.

Answer 1

[Stephen23]

Part One: Dynamic Fieldnames

This is called dynamic fieldnames, and the syntax is pretty simple: you just need to add some parentheses:

Data.(str)

where str is a variable that defines the name of that field. In your case you could generate the field name using sprintf:

tmp = Data.(sprintf('f%d',x));
plot(tmp(:,1),tmp(:,2)

Part Two: Using Non-Scalar Structures Instead

Although there might be a good reason for your naming, these sequentially-numbered names might be better replaced by a non-scalar structure. These would allow you to simply call each element without needing to generate the fieldname dynamically:

tmp = Data(x).field;
plot(tmp(:,1),tmp(:,2)

which you can see is much simpler to work with! This also has the advantage that you can do neat tricks like this:

Data.field

returns a list of all field values, in a comma-separated list. This means, for example, that you can create a cell array of all field values simply by doing this:

{Data.field}

or concatenate them into one array like this:

[Data.field]

Part Three: Use a Cell/Numeric Array Instead

If this is the only data that is being stored in the structure, then using a structure to store it is very large hammer to crack this nut with: it would be simpler to use either:

• one cell array, and each cell contains one matrix
• one simple numeric array, using the dimensions to distinguish the semantic meaning (e.g. the third dimension corresponds to different data measurements).