function [idx rows_idx] = anyRows(db, rows)
% anyRows - Returns db rows matching any of the given rows.
%
% Usage:
% idx = anyRows(db, rows)
%
% Description:
% The db rows are compared to each row and row indices succeeding any of
% these comparisons are returned.
%
% Parameters:
% db: A tests_db object.
% rows: Row array, matrix or database to be compared with db rows.
%
% Returns:
% idx: A logical column vector of matching db row indices.
% rows_idx: Indices of rows entries corresponding to each db
% row. Non-matching entries were left as NaN.
%
% Example:
% >> db(anyRows(db(:, 'trial'), [12; 46; 37]), :)
% returns a db with rows having trial equal to any of the given values.
%
% See also: compareRows, eq, tests_db
%
% $Id$
%
% Author: Cengiz Gunay <cgunay@emory.edu>, 2004/09/17
% Copyright (c) 2007 Cengiz Gunay <cengique@users.sf.net>.
% This work is licensed under the Academic Free License ("AFL")
% v. 3.0. To view a copy of this license, please look at the COPYING
% file distributed with this software or visit
% http://opensource.org/licenses/afl-3.0.php.
% TODO: optimize by using sort
% Compare two dbs?
if isa(rows, 'tests_db')
rows = rows.data;
end
% sanity check
if dbsize(db, 2) ~= size(rows, 2)
error(['Rows contains ' num2str(size(rows, 2)) ' columns, but db has ' ...
num2str(dbsize(db, 2)) ' columns. They must match for comparison.']);
end
% prepare variables for faster processing
num_rows = size(rows, 1);
num_db_rows = size(db.data, 1);
ones_matx = ones(num_db_rows, 1);
idx = false(num_db_rows, 1);
rows_idx = repmat(NaN, num_db_rows, 1);
db_data = db.data;
% DISABLED: look for optimization.
% unfortunately this is not even faster due to the memory requirement :(
if false && dbsize(db, 2) == 1 && num_db_rows * num_rows < 10000000
idx = all(abs(( db.data * ones(1, num_rows) - (rows * ones_matx')')) <= ...
eps(0), 2);
else
% Calculate multiple rows by tail recursion
for row_num=1:num_rows
% Find doesn't work in two dimension comparisons
% Thus, use algorithm:
% - duplicate row to a matrix of same size with db
% - subtract from db
matching_db_rows = ...
all(abs(db_data - (ones_matx * rows(row_num, :))) <= eps(0), 2);
rows_idx(matching_db_rows) = row_num;
idx = idx | matching_db_rows;
end
end
end