How to skip error and continue (NOT in loop)?

If I have the following code, how to display x(10) after x(9999) causing an error?

x = ones(1, 10);
try
    x(1)
    x(9999)
    x(10)
catch ME
    fprintf(ME.message)
end

The output I got is:

ans =

     1

Index exceeds the number of array elements (10).

Thanks!

Answer 1

[Walter Roberson]

x = ones(1,10);
idx = [1 9999 10];
for K = 1 : length(idx)
    try
        x(K)
    catch ME
      fprintf(ME.message)
    end
end

[n33]

Thanks Walter, this works good.

So I assume if I have several user-defined functions, the only way to guarantee all functions get executed in a single run is to have several independent try-catch blocks or somehow put those functions in a loop?

%% GOOD
try
    myfun1()
catch ME
end

try
    myfun2()
catch ME
end

%% BAD
try
    myfun1()
    myfun2()
catch ME
end

[Walter Roberson]

funcs = {@myfun1, @myfun2};
for K = 1 : length(funcs)
    try
      funcs{K}();
    catch ME
        fprintf(ME.message);
    end
end

[Fernando]

What is the part of "not in a loop" you didn't understand?

[Walter Roberson]

@Fernando

I'm not sure which part of "not in a loop" I did not understand. Perhaps you could explain?

[Fernando]

if one has commands that truly cannot be done in a loop, then how can you skip a command error without having to try/catch every single command?

say:

A = command1thatcanpossiblycauseanerror;
B = command2thatcanpossiblycauseanerror;
C = command3thatcanpossiblycauseanerror;
D = command4thatcanpossiblycauseanerror;
E = command5thatcanpossiblycauseanerror;
F = command6thatcanpossiblycauseanerror;
G = command7thatcanpossiblycauseanerror;
H = command8thatcanpossiblycauseanerror;
I = command9thatcanpossiblycauseanerror;
J = command10thatcanpossiblycauseanerror;
K = command11thatcanpossiblycauseanerror;
L = command12thatcanpossiblycauseanerror;
M = command13thatcanpossiblycauseanerror;
N = command14thatcanpossiblycauseanerror;
O = command15thatcanpossiblycauseanerror;
P = command16thatcanpossiblycauseanerror;
Q = command17thatcanpossiblycauseanerror;
R = command18thatcanpossiblycauseanerror;
S = command19thatcanpossiblycauseanerror;
T = command20thatcanpossiblycauseanerror;
U = command21thatcanpossiblycauseanerror;
V = command22thatcanpossiblycauseanerror;
W = command23thatcanpossiblycauseanerror;
X = command24thatcanpossiblycauseanerror;
Y = command25thatcanpossiblycauseanerror;
Z = command26thatcanpossiblycauseanerror;

%where each commandXXthatcanpossiblycauseanerror is some command, really
%anything, something that cannot possibly be in a for loop.

%as the author asked.

[Walter Roberson]

as the author asked

Sorry, could you explain to me where the author of the original question asked that?

[Steven Lord]

A = command1thatcanpossiblycauseanerror;
B = command2thatcanpossiblycauseanerror;
C = command3thatcanpossiblycauseanerror;
D = command4thatcanpossiblycauseanerror;
% etc

That has a similar code smell to the frequently asked question about dynamically creating variables with numbered names like x1, x2, x3, etc., for which the general consensus is to avoid doing that.

Rather than posing a hypothetical scenario, can you share a little more detail about your actual application where you use a pattern like that code? What do those commands that can possibly throw an error do? Are they independent or are the outputs of one or more of those functions the inputs to later of those functions?

[Walter Roberson]

I am having difficulty thinking of any code that is such that

A = command1thatcanpossiblycauseanerror;

is valid code, but

for dfpSzcGKKqyMbgIcvYXqkicEiRRKtIUpb_vjDshBQupYrSfhm_GS_xNidaex = 1 : 1
    A = command1thatcanpossiblycauseanerror;
end

"cannot" work.

The only things I can come up with are along the lines of "needing" dbstack to be able to pinpoint the function name... ummm, no, even that would not explain things.