For completeness, here is an excerpt from the MATLAB help file on `unique`:
[C <http://se.mathworks.com/help/matlab/ref/unique.html#outputarg_C>,ia <http://se.mathworks.com/help/matlab/ref/unique.html#outputarg_ia>,ic <http://se.mathworks.com/help/matlab/ref/unique.html#outputarg_ic>] = unique(A <http://se.mathworks.com/help/matlab/ref/unique.html#inputarg_A>) also returns index vectors ia and ic. - If A is a vector, then C = A(ia) and A = C(ic). - If A is a matrix or array, then C = A(ia) and A(:) = C(ic). - If A is a table, then C = A(ia,:) and A = C(ic,:). Which of the cases is relevant for you? // T On Tuesday, April 21, 2015 at 11:16:12 AM UTC+2, Stephan Buchert wrote: > > In Matlab I can do, for example, with "daynum" fractional day numbers: > > idaynum = floor(daynum); > [umjd, im, iu] = unique(idaynum); % <= how do I get the "iu" in Julia? > > % I need an observation "pms" listed per day in a file. > % For performance reasons the parameter is retrieved from the file only > once for each of the different integer days > upms = pi*umjd; % <= for illustration only, instead of really > retrieving the observed parameter > pms = upms(iu); > % Now "pms" matches "daynum", ie. when daynum=[1.0, 1.5, 1.8, 2.0, 2.2, > 3.1], then pms approx. [3.14, 3.14, 3.14, 6.28, 6.28, 9.42] > > Probably this can be done elegantly also in Julia, but, coming from > Matlab, for me not easy to see how? >
