mean.m: Accept large DIM inputs and calculate Matlab-compatible output (bug #63411) * mean.m: Accept large DIM inputs by removing any dimensions larger than the dimensions of the input array. Use vectorized operations, rather than for loop, for this input validation. Pre-declare output array before for loop for performance. Use ipermute() after calculations to return correctly-dimensioned result (checked with Matlab). Change BIST tests results to match correct behavior.

((nil .
      ((c-file-style . "gnu")
       (indent-tabs-mode . nil)
       (fill-column . 72)))
 (change-log-mode . ((indent-tabs-mode . t)))
 (makefile-mode . ((indent-tabs-mode . t))))