Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

New to MATLAB?

An easy Array question

Asked by Hello kity

Hello kity (view profile)

on 12 Dec 2012

I have a set of data seprated by Nans, example; DATA= [1; 2; 3; 4; NaN ; 6 ;7 ; NaN]

I want to do a calculation in every set of rows (example 1- 4 and 6-7).

So I programmed and found all the begin and end indices of each row (in this example, begin [ 1 ; 4] and end [6 ; 7]).

so this is what i want: Have two columns with begin and end indices;

column 1;  A=[1 ; 6 ] %%%these are the begin indices
column 2;  B[ 4 ;7 ] %% these are the end indices

now i want created a new array (call it new array) that has counted 1-4 and 6-7.

Then all i need to do is out=data(newarray) and i ll have my rows which i use for calc

0 Comments

Hello kity

Hello kity (view profile)

Products

No products are associated with this question.

2 Answers

Answer by Sean de Wolski

Sean de Wolski (view profile)

on 12 Dec 2012

Are you sure you need a cell array?

I would just use a for-loop and loop over the start and end indices:

sidx = start index
eidx = end index
for ii = 1:numel(eidx)
 do_something_with(X(sidx(ii):eidx(ii)));
end

Of course this is also how I would build the cell array (i.e. replace

do_something_with with C(idx) = {}

And preallocate C with

C = cell(numel(sidx,1));

1 Comment

Hello kity

Hello kity (view profile)

on 12 Dec 2012

i changed my question a bit, maybe it is more clear now.

Sean de Wolski

Sean de Wolski (view profile)

Answer by Andrei Bobrov

Andrei Bobrov (view profile)

on 13 Dec 2012
Edited by Andrei Bobrov

Andrei Bobrov (view profile)

on 13 Dec 2012
DATA= [1; 2; 3; 4; NaN ; 6 ;7 ; NaN];
a = [1;4];
b = [6;7];
d = DATA(~isnan(DATA));
out = mat2cell(d(:),[diff(a)+1,diff(b)+1],1);

2 Comments

Jan Simon

Jan Simon (view profile)

on 13 Dec 2012

I think there must be a ~ before isnan.

Andrei Bobrov

Andrei Bobrov (view profile)

on 13 Dec 2012

Hi Jan, thank you. Corrected.

Andrei Bobrov

Andrei Bobrov (view profile)

Contact us