arrayfun doesn't work with rmoutliers

3 Ansichten (letzte 30 Tage)
Tom K.
Tom K. am 5 Jan. 2022
Bearbeitet: Tom K. am 5 Jan. 2022
Goal
Remove outliers along dimension d of a 3D matrix using arrayfun and rmoutliers (or superior alternatives).
Minimal Working Example
Take the following matrix:
A = [0 2 0 0 0; 0 0 0 4 0; 3 0 0 3 0];
A(:,:,2) = A
A =
A(:,:,1) = 0 2 0 0 0 0 0 0 4 0 3 0 0 3 0 A(:,:,2) = 0 2 0 0 0 0 0 0 4 0 3 0 0 3 0
Since each 1D array along dimension d will have a different number of outliers, the output of arrayfun will have to be a cell array (by specifying "UniformOutput" = false). Calling rmoutliers through arrayfun returns the original matrix with the outliers (undesired behaviour):
arrayfun(@rmoutliers,A,"UniformOutput",false)
ans = 3×5×2 cell array
ans(:,:,1) = {[0]} {[2]} {[0]} {[0]} {[0]} {[0]} {[0]} {[0]} {[4]} {[0]} {[3]} {[0]} {[0]} {[3]} {[0]} ans(:,:,2) = {[0]} {[2]} {[0]} {[0]} {[0]} {[0]} {[0]} {[0]} {[4]} {[0]} {[3]} {[0]} {[0]} {[3]} {[0]}
The function works as expected when called on each row by itself:
rmoutliers(A(1,:,1)) % removes the 2
ans = 1×4
0 0 0 0
rmoutliers(A(3,:,1)) % removes both 3's
ans = 1×3
0 0 0
Hopefully I'm missing something trivial.

Akzeptierte Antwort

Voss
Voss am 5 Jan. 2022
A = [0 2 0 0 0; 0 0 0 4 0; 3 0 0 3 0];
A(:,:,2) = A
A =
A(:,:,1) = 0 2 0 0 0 0 0 0 4 0 3 0 0 3 0 A(:,:,2) = 0 2 0 0 0 0 0 0 4 0 3 0 0 3 0
C = num2cell(A,2)
C = 3×1×2 cell array
C(:,:,1) = {[0 2 0 0 0]} {[0 0 0 4 0]} {[3 0 0 3 0]} C(:,:,2) = {[0 2 0 0 0]} {[0 0 0 4 0]} {[3 0 0 3 0]}
cellfun(@rmoutliers,C,"UniformOutput",false)
ans = 3×1×2 cell array
ans(:,:,1) = {[0 0 0 0]} {[0 0 0 0]} {[ 0 0 0]} ans(:,:,2) = {[0 0 0 0]} {[0 0 0 0]} {[ 0 0 0]}

Weitere Antworten (1)

Mike Croucher
Mike Croucher am 5 Jan. 2022
arrayfun operates on every element of the array. So
arrayfun(@rmoutliers,A,"UniformOutput",false)
is like doing
rmoutliers(0)
rmoutliers(0)
rmoutliers(3)
etc
  1 Kommentar
Tom K.
Tom K. am 5 Jan. 2022
Bearbeitet: Tom K. am 5 Jan. 2022
Hi Mike,
I was able to produce the desired results using @Benjamin's solution.
Thanks!

Melden Sie sich an, um zu kommentieren.

Kategorien

Mehr zu Matrix Indexing finden Sie in Help Center und File Exchange

Produkte


Version

R2021a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by