I'm trying to port a working MATLAB code in Python. I am trying to create the var array with size (rows, cols). If an exception is raised, I catch it and try again to create the var array with size (rows, cols-1). If cols gets to zero, then I cannot do anything else, so I rethrow the previously caught exception.

The code snippet is as follows:

% rows, cols init
success = false;
while(~success)
    try
        var = zeros(rows, cols);
        success = True;
    catch ME
        warning('process memory','Decreasing cols value because out of memory');
        success = false;
        var = [];
        cols = cols - 1;

        if (cols < 1)
            rethrow(ME);
        end
    end
end

The doc of rethrow states:

Instead of creating the stack from where MATLAB executes the method, rethrow preserves the original exception information and enables you to retrace the source of the original error.

My question is: What should I write in Python to have the same results with MATLAB's rethrow?

In Python, I wrote the following. Is that sufficient?

# rows, cols init
success = False
while not success:
    try:
        var = np.zeros([rows, cols])
        success = True
    except MemoryError as e:
        print('Process memory: Decreasing cols value because out of memory')
        success = False
        var = []
        cols -= 1

        if cols < 1:
            raise(e)

CodePudding user response：

The corresponding syntax is just raise: raise e (note: it’s not a function) adds a stacktrace entry for itself. (In Python 2, it replaced the previous stacktrace like MATLAB’s ordinary throw, but Python 3 extends it.)