how do i find argmax?
451 Ansichten (letzte 30 Tage)
Ältere Kommentare anzeigen
shikhar
am 21 Mai 2013
Bearbeitet: James Tursa
am 18 Mai 2021
how do i find argmax of, say x=[1 2] without any user defined function?
0 Kommentare
Akzeptierte Antwort
Walter Roberson
am 21 Mai 2013
[argvalue, argmin] = min(x);
[argvalue, argmax] = max(x);
1 Kommentar
Weitere Antworten (2)
Lukasz Laszko
am 28 Okt. 2019
Bearbeitet: Lukasz Laszko
am 28 Okt. 2019
I'm afraid the answer by Walter Roberson is not correct. According to docs the second returned value contains indices of the minimum/maximum values of x. So to find argmax, you need to call x(max_ind),
eg.
x=0.1:.01:4*pi;
[max_val max_ind]=max(cos(x))
argmax=x(max_ind)
-->> max_val=1
-->> argmax=2pi
-->> max_ind=619
3 Kommentare
James Tursa
am 17 Mai 2021
Bearbeitet: James Tursa
am 18 Mai 2021
@Lukasz Laszko Argmax is defined as the elements of the domain (i.e., in this case the indexes) where the function (i.e., in this case the vector) obtains the maximum. See these definitions:
Walter's answer is correct according to these definitions.
Your use of cos(x) introduces another function inbetween that was not part of the original question. Did the original question have cos(x) that was subsequently edited out?
Walter Roberson
am 18 Mai 2021
In the case a function that is effectively f(x(K)) and what you wanted was the x(K) then it would not be unreasonable to take
fx = f(x);
[~, argidx] = max(fx);
argmax = x(argidx);
This adds an extra level of indirection, but it would be a common one.
Siehe auch
Kategorien
Mehr zu GPU Computing finden Sie in Help Center und File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!