How to insert data to a matrix based on index values stored in a matrix?

Hello!

I am a beginner in Matlab and would like to ask a question regarding inserting NaN- values to a matrix according to index values that are stored in a matrix.

I have the following case:

a is a matrix containing data I want to insert NaN- values to the start and end of the vector.

a = vector 250x111

b is a matrix where first row has index values of the start cut-off point and second row information of end cut-off point. For example start cut-off point could be 20 and end cut-off point 105 for a certain data point.

b = matrix 2x111

I believe I can implement this with following by using for- loop through data (example for the first data point):

a( 1:b(1,1) ) = NaN;

a( b(2,1):end) = NaN;

However, I am wondering if there is a way or function to conduct this smoothly? I have noticed while reading this forum that there is usually more efficient ways than using for- loops to conduct this kind of operations and there for wanted to ask for tips.

Thank you already in advance!

[Stephen23]

As you have just one vector, surely this is just:

a(1:max(b(1,:))) = NaN;
a(min(b(2,:)):end) = NaN;

(note that in your example you have swapped the row and column indexing).

[Martti Ilvesmäki]

Thank you for the answer, and true that my example indexing was indeed accidentally swapped.

I just realized I wrote incorrectly description of vector a. It is actually a matrix containing 111 vectors of length 250.

a = matrix 250x111

b contains cut-off positions for start and end for each vector as described in the original message.

I have edited the post accordingly now.

Answer 1

[Stephen23]

a = randi(9,7,5)

a = 7×5

8 9 6 7 6 4 7 9 8 3 8 9 4 1 3 7 8 5 6 8 9 6 3 3 9 3 2 4 4 6 7 8 3 3 8

b = [2,2,3,1,2;6,5,7,6,5]

b = 2×5

2 2 3 1 2 6 5 7 6 5

rwv = 1:size(a,1);
idx = rwv(:)<=b(1,:) | rwv(:)>=b(2,:);
a(idx) = NaN

a = 7×5

NaN NaN NaN NaN NaN NaN NaN NaN 8 NaN 8 9 NaN 1 3 7 8 5 6 8 9 NaN 3 3 NaN NaN NaN 4 NaN NaN NaN NaN NaN NaN NaN

If you are using a MATLAB version prior to R2016b replace both of the logical comparisons with BSXFUN and function handles to the same logical functions.

[Martti Ilvesmäki]

Thank you, this worked very well and answered my question!