How to use interp2 to resize arrays of different sizes?

Hello,

I have an excel sheet test data with results in individual sheets. I am plotting two columns in each sheets. But the array sizes are not exactly the same in all sheets. For example: - In sheet 1- My test datas are of the size 1202x2 - In sheet 2- My test datas are of the size 1752x2 - In sheet 3- My test datas are of the size 1402x2 How can I use interp2 to resize the bigger ones to the size of 1202x2. I was able to do with interp1 individually but it would be better if I implement interp2?

Thanks a lot in advance.

[Adam]

You should just be able to do it similarly to interp1, but set the output size the same as the input size in the y direction

Answer 1

[dpb]

Not sure it will actually be any faster than using interp1 by column for only two columns and will certainly be far more memory intensive to build the grid interp2 needs but the technique is

targetSize = size(Sheet1Vector);  % would be 1202x2 here
sourceSize = size(SheetNVector);  % 1752x2 for second series 
[X,Y] = meshgrid(linspace(1,sourceSize(2),targetSize(2)), ...
                 linspace(1,sourceSize(1),targetSize(1)));  % build the interpolant grid
New2DArray=interp2(x,X,Y);   % interpolate each x in turn

For just an easy example to see...use a small test dataset

>> x=rand(12,2);
>> ts=[10 2];
>> [X Y]=meshgrid(linspace(1,2,2),linspace(1,12,10))
X =
   1     2
   1     2
   1     2
   1     2
   1     2
   1     2
   1     2
   1     2
   1     2
   1     2
Y =
  1.0000    1.0000
  2.2222    2.2222
  3.4444    3.4444
  4.6667    4.6667
  5.8889    5.8889
  7.1111    7.1111
  8.3333    8.3333
  9.5556    9.5556
 10.7778   10.7778
 12.0000   12.0000
>> interp2(x,X,Y)
ans =
  0.9836    0.6382
  0.9138    0.2984
  0.4237    0.6975
  0.6539    0.7978
  0.1446    0.7331
  0.9451    0.5553
  0.5568    0.1249
  0.2970    0.1182
  0.8009    0.3741
  0.0561    0.0216
>> whos ans
Name       Size            Bytes  Class     Attributes

ans       10x2               160  double

For this special case, you can easily build the X, Y arrays by inspection and only Y will change from sheet to sheet; just showed the general form where both X and Y could be variable.

There's a function imresize in the Image Processing toolbox that does this in one call if you happen to have that...

[Swaminath VENKATESWARAN]

Hello,

Many thanks for your support. It was really useful. But the thing was it took some time. Actually I found another way using interp1. I assigned my required x-y values to be resized into cells. I extract the cell which has the least size and move it to the last position, take it as reference and with the help of 'for' loop I implement interp1 on the other cells. It looks bit lengthy but was very effective. Thanks a lot for your support again.