using sum() with a vector and increment

4 views (last 30 days)
Hi peeps, sorry about this rather simple question; I'm very new to this and cannot wrap my head around it. Dummy-Code:
A= [1 2 3 4 5; 6 7 8 9 10; 11 12 13 14 15]; %create 3x5 matrix
a = [2 3 4]; %row vector
b = A(1,:); %extract row 1 from Matrix A
c= b(a+1); %create vector from positions+1 defined with a
d=[1 c]; %create vector with 1 and positions
dsum=sum(A(1,d:d+1),1); %??????
Why does d:d+1 give me the first two positions of row 1 (and 2 and 3 if I change it so A(2,d:d+1))? I understand that : can be used to create vectors with certain increments, but it simply makes no sense to me. I'm sure there is a logical explanation. Anybody? Help is greatly appreciated.

Accepted Answer

Jan
Jan on 11 Apr 2018
Edited: Jan on 11 Apr 2018
The colon operator ":" needs scalars as input. Then operation "vector:vector" is not defined. Unfortunately Matlab does not stop with an error, if you provide vectors as input, but only the first element is used.
v = [7,8,9]
v:v+1 % This is the same as: 7:7+1
% The rest of v is ignored
In your case only the first element 1 is used, such that "d:d+1" is the same as "1:2".
This behavior is a typical source of unexpected behavior in:
for k = 1:size(v)
when size(v,2) or numel(v) is meant.
  1 Comment
Guillaume
Guillaume on 11 Apr 2018
Then operation "vector:vector" is not defined
I disagree, it is very well defined in that it is explicitly documented:
If you specify nonscalar arrays, then MATLAB interprets j:i:k as j(1):i(1):k(1).
However, yes, it would probably make more sense if matlab simply errored.

Sign in to comment.

More Answers (1)

Isabella Kratzer
Isabella Kratzer on 11 Apr 2018
Thank you both Jan and Guillaume!

Categories

Find more on Creating and Concatenating Matrices in Help Center and File Exchange

Community Treasure Hunt

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

Start Hunting!