Convert string to variable.

I'm pulling a string from a table 'Matrix1'. Before that, I have defined Matrix1 as an actual matrix. What I want to do is pull the string from the table and have matlab recognize the name and use the matrix. I have to pull it from the table, because the table is changing and may sometimes say 'Matrix2' or 'Matrix3'.

So what I have now is a table (named 'Table') and a column (named 'Territory'). I'm using the command Table.Territory(3) to access the third territory in the list. But it comes through as a string and I need it to come through as a matrix. I have already saved that territory name as a matrix.

[Steven Lord]

Can you dynamically create and reference variables with numbered names like x1, x2, x3, etc.? Yes.

Should you do this? The general consensus is no. That Answers post explains why this is generally discouraged and offers several alternative approaches.

[DKonk]

I'm not entirely sure what a dynamic variable is, but I'm fairly sure I dont want to do this. So I have an excel doc that will be changing frequently (maybe this makes it dynamic?). In my code I will have:

US_matrix = [1 0 0; 0 1 0]

UK_matrix = [0 0 1; 1 0 0] % Just example matrices

Now I need to pull a string from the table, either 'US_matrix' or 'UK_matrix' and access the respective matrix.

[Stephen23]

"I'm not entirely sure what a dynamic variable is, but I'm fairly sure I dont want to do this."

Yet it seems to be exactly what you are attempting. Which would force you into writing slow, complex, inefficient, obfuscated, insecure, buggy code which is hard to debug... just to access your data:

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

Best avoided.

"So I have an excel doc that will be changing frequently (maybe this makes it dynamic?)"

No, the fact your file can change is unrelated.

Answer 1

[Stephen23]

This is the cause of your problems right here:

US_matrix = [1 0 0; 0 1 0]

UK_matrix = [0 0 1; 1 0 0]

Having meta-data in variable names is a big sign that you are doing something wrong. Remember, meta-data is data, and data belongs in variables (i.e. arrays) and not in variable names.

As soon as you define lots of separate variables like that, each with their own name, then you have really painted yourself into a dark corner: then you must write slow, complex, inefficient code just to access your variable names dynamically. Ouch. The simple, very efficient, and recommended solution is to use arrays and indexing to access your data (because that is how MATLAB works), for example here are some much better approaches:

Method one: cell array and indexing (recommended):

want = 'UK'; % you get this from the Excel file, or wherever.
data = {[1,0,0;0,1,1],[0,0,1;1,0,0]};
name = {         'US',          'UK'};
idx = strcmpi(name,want);
data{idx} % this is how MATLAB works efficiently, using indexing. Not like your approach.

ans = 2×3

0 0 1 1 0 0

Method two: non-scalar structure (recommended):

A(1).name = 'US';
A(1).data = [1,0,0;0,1,1];
A(2).name = 'UK';
A(2).data = [0,0,1;1,0,0];
idx = strcmpi({A.name},want);
A(idx).data

ans = 2×3

0 0 1 1 0 0

Method three: scalar structure with dynamic fieldnames (still a bit fragile, probably best avoided):

B.US = [1 0 0; 0 1 0];
B.UK = [0 0 1; 1 0 0];
B.(want) % dynamic fieldname

ans = 2×3

0 0 1 1 0 0

https://www.mathworks.com/help/matlab/matlab_prog/generate-field-names-from-variables.html

Note:

• Some of these could even be vectorized very easily (e.g. ISMEMBER).
• You could do something similar using tables: https://www.mathworks.com/help/matlab/tables.html
• Remember that meta-data is data (and shoud not be part of a variable name).

[DKonk]

Method 1 works like a charm. You helped a lot and simplified my code. Much appreciated!