Skip to content

randsample.m: fix categorical vector support and add BISTs #346

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
SergeantQuickscoper wants to merge 2 commits into
base: main
Choose a base branch
from
Open

randsample.m: fix categorical vector support and add BISTs #346

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7693a17
randsample.m: tiny doc fix as strcat trims trailing whitespace
SergeantQuickscoper Dec 24, 2025
8efe7a5
randsample.m: fix support for categorical vectors, add vector test ca…
SergeantQuickscoper Dec 24, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
92 changes: 82 additions & 10 deletions inst/randsample.m
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
## Copyright (C) 2014 - Nir Krakauer
## Copyright (C) 2022 Andreas Bertsatos <abertsatos@biol.uoa.gr>
## Copyright (C) 2025 Andreas Bertsatos <abertsatos@biol.uoa.gr>
##
## This file is part of the statistics package for GNU Octave.
##
Expand All @@ -24,7 +24,7 @@
## Sample elements from a vector.
##
## Returns @var{k} random elements from a vector @var{v} with @var{n} elements,
## sampled without or with @var{replacement}.
## sampled without or with @var{replacement}, with an optional weight vector.
##
## If @var{v} is a scalar, samples from 1:@var{v}.
##
Expand All @@ -38,26 +38,26 @@
## @seealso{datasample, randperm}
## @end deftypefn

function y = randsample (v, k, replacement=false ,w=[])
function y = randsample (v, k, replacement=false, w=[])

if (isscalar (v) && isreal (v))
n = v;
vector_v = false;
elseif (isvector (v))
n = numel (v);
n = length (v);
vector_v = true;
else
error ("randsample: The input v must be a vector or positive integer.");
endif

if k < 0 || ( k > n && !replacement )
error (strcat ("randsample: The input k must be a non-negative ", ...
"integer. Sampling without replacement needs k <= n."));
error (strcat ("randsample: The input k must be a non-negative", ...
" integer. Sampling without replacement needs k <= n."));
endif

if (all (length (w) != [0, n]))
error ("randsample: the size w (%d) must match the first argument (%d)", ...
length(w), n);
length (w), n);
endif


Expand All @@ -78,8 +78,8 @@
if (! any (Idup))
break
else
Idup(idx) = Idup; # find duplicates in original vector
w(y) = 0; # don't permit resampling
Idup (idx) = Idup; # find duplicates in original vector
w (y) = 0; # don't permit resampling
## remove duplicates, then sample again
y = [y(! Idup), (weighted_replacement (sum (Idup), w))];
endif
Expand All @@ -95,7 +95,7 @@

function y = weighted_replacement (k, w)
w = w / sum (w);
w = [0, cumsum(w(:))'];
w = [0, (cumsum (w(:))')];
## distribute k uniform random deviates based on the given weighting
y = arrayfun (@(x) find (w <= x, 1, "last"), rand (1, k));
endfunction
Expand Down Expand Up @@ -135,3 +135,75 @@
%! assert (size(x), [k 1]);
%! x = randsample(k, k, false, 1:k);
%! assert (size(x), [1 k]);

%!test
%! n = 20;
%! k = 5;
%! p = 1:n;
%! x = randsample(p, k);
%! assert (isnumeric(x));
%! assert (size(x), [1 k]);
%! x = randsample(p, k, true);
%! assert (isnumeric(x));
%! assert (size(x), [1 k]);
%! x = randsample(p, k, false);
%! assert (isnumeric(x));
%! assert (size(x), [1 k]);
%! k = 30;
%! x = randsample(p, k, true);
%! assert (isnumeric(x));
%! assert (size(x), [1 k]);

%!test
%! p = categorical({'a', 'b', 'c', 'd', 'a'});
%! k = 3;
%! x = randsample(p, k, true);
%! assert (iscategorical(x));
%! assert (size(x), [1 k]);
%! x = randsample(p, k, false);
%! assert (iscategorical(x));
%! assert (size(x), [1 k]);
%! k = 30;
%! x = randsample(p, k, true, ones(length(p),1));
%! assert (iscategorical(x));
%! assert (size(x), [1 k]);

%!test
%! p = {'a', 'b', 'c', 'd', 'a'};
%! k = 2;
%! x = randsample(p, k, true);
%! assert (iscell(x));
%! assert (size(x), [1 k]);
%! x = randsample(p, k, false);
%! assert (iscell(x));
%! assert (size(x), [1 k]);
%! k = 30;
%! x = randsample(p, k, true, ones(length(p),1));
%! assert (iscell(x));
%! assert (size(x), [1 k]);

%!test
%! p = string({'a', 'b', 'c', 'd', 'a'});
%! k = 2;
%! x = randsample(p, k, true);
%! assert (isstring(x));
%! assert (size(x), [1 k]);
%! x = randsample(p, k, false);
%! assert (isstring(x));
%! assert (size(x), [1 k]);
%! k = 30;
%! x = randsample(p, k, true, ones(length(p),1));
%! assert (isstring(x));
%! assert (size(x), [1 k]);

%!error <randsample: The input v must be a vector or positive integer.> ...
%! randsample ([1 2 3; 1 2 3], 5)

%!error <randsample: The input k must be a non-negative integer.> ...
%! randsample (10, -1)

%!error <Sampling without replacement needs k <= n.> ...
%! randsample (10, 100)

%!error <randsample: the size w .* must match the first argument .*> ...
%! randsample (10, 5, false, ones(5,1))