STRPAT searches for EXACTLY matching patterns
in a character or numeric string and replaces
the occurrences with another pattern.

a single or double precision pattern
may include NaNs

% examples

% - character strings

% note difference
% STRREP
strrep('aa_a_aaa_aa','a','XXX')
% XXXXXX_XXX_XXXXXXXXX_XXXXXX
%
% STRPAT
strpat('aa_a_aaa_aa','a','XXX')
% aa_XXX_aaa_aa

strpat('aa_a_aaa_aa','aa','X')
% X_a_aaa_X

% - numeric strings
strpat([1:3,pi,5:7],pi,nan)
% 1 2 3 NaN 5 6 7

strpat(pi*(1:6),pi*(4:5),[])
% 3.1416 6.2832 9.4248 18.85

strpat(pi*(1:6),pi*(4:5),[nan inf -inf nan])
% 3.1416 6.2832 9.4248 NaN Inf -Inf NaN 18.85

strpat(pi*(1:6),pi*(4:5),1:3)
% 3.1416 6.2832 9.4248 1 2 3 18.85

% - PATTERN including NaNs
strpat([-10,nan,1,-10,nan,nan,1,nan,1,-10],[nan,1],inf)
% -10 Inf -10 NaN Inf Inf -10

see also
>> help strpat
or the screenshot for help on this self-explanatory function