Find adjacent values of a vector corresponding to another vector

Hi, I am trying to find the adjacent points of a 1-D vector corresponding to the elements of another column vector. It is very hard to explain. I will explain it from an example.

A=[ 11; 22; 32; 44; 51;]; B=[11.4 32.4 36.6]; So I need a matrix with the adjacent points of each element of the vector B of vector A as in, Ans1=[11 22; 32 44; 32 44;] also the corresponding locations as in, ans2[1 2; 3 4; 3 4;] Appreciate any help.

Answer 1

[Matt J]

idx=discretize(B,A); 
Ans2=[idx(:),idx(:)+1]; 
Ans1=A(Ans2);

[anton fernando]

Thanks. :)

[anton fernando]

Your answer worked for my example. But it doesn't work for my data set. Let me get you my data set.

A=[1
1
0.71100003
0.62000000
0.53899997
0.46700001
0.40200001
0.34500000
0.29499999
0.25299999
0.21699999
0.18600000
0.16000000
0.13699999
0.11700000
0.099799998
0.085000001
0.072499998
0.061799999
0.052600000
0.044700000
0.037999999
0.032299999
0.027400000
0.023200000
0.019700000
0.016700000
0.014100000
0.011900000
0.0099799996
0.0083999997
0.0070600002
0.0059400001
0.0049999999
0.0042099999
0.0035500000
0.0029899999
0.0025200001
0.0021200001
0.0018000000
0.0015200000
0.0012900000
0.0011000000
0.00094200001
0.00080699997
0.00069299998
0.00059700001
0.00051400001
0.00044300000
0.00038200000
0.00033000001
0.00028599999
0.00024800000
0.00021499999
0.00018800001
0.00016400000
0.00014300000
0.00012600000
0.00011000000
9.6999996e-05];
B=[0.986923267000000
0.672382994000000
0.458089201000000
0.312092540000000
0.212626172000000
0.144860520000000
0.0986923270000000
0.0672382990000000
0.0458089200000000
0.0312092540000000
0.0212626170000000
0.0144860520000000
0.00986923300000000
0.00672383000000000
0.00458089200000000
0.00312092500000000
0.00212626200000000
0.00144860520000000
0.000986923000000000
0.000672383000000000
0.000458089000000000
0.000312093000000000
0.000212626000000000
0.000144860000000000
9.86923000000000e-05];

Thank you in advance.

[Matt J]

This A data is unsorted. Your problem is ill-defined for unsorted A. What would "adjacency" mean in that case?

[anton fernando]

Yes. Because these are pressure levels in the atmosphere. Can I do it without sorting it? Adjacent means the values nearest or closet. If you find a value in between two values I call those two values adjacent.

[Matt J]

You should modify your posted question with a similarly small example with unsorted A. How can a value of B fall "in between two values" of A if A is not sorted?

[anton fernando]

I am sorry. I meant by not sorted that there can be NaN values and repeated values in the vector. But initially I don't have to worry about it. But sure, I can flip the matrix and make it monotonically increasing. But vector A is monotonically decreasing.

[anton fernando]

Once you flipped the matrices it works. Thanks. :)

Answer 2

[Jos (10584)]

Here is a solution that finds the element in A that is just before or just after each element in B.

A=[ 11; 22; 32; 44; 51;]; B=[11.4 32.4 36.6]
clear ans2
ans2(:,1) = nearestpoint(B, A, 'before') ;
ans2(:,2) = nearestpoint(B, A, 'after') ;
ans1 = A(ans2) % will fail when elements of B do not fall between elements of A. Check for that!

My NEARESTPOINT utility can be downloaded from the File Exchange:

https://uk.mathworks.com/matlabcentral/fileexchange/8939-nearestpoint-x--y--m-