Stretch downsampled data to fit on same x axis

I have an array1 sampled at 2Khz for 500 seconds.

I want to visualize the hilbert transform of the data at a lower sampling rate (i.e. smoother trend)

To accomplish this I used array2 = downsample(array1, 4) then performed the hilbert transform.

As a result, array2 is shorter than array1 even though they represent the same amount of time.

How can I now "stretch" array2 back to the length of array1 so that I can view them on a plot simultaneously with the same time axis?

Fs = 2000
t = 500
>> ar1 = 1:1:Fs*t;
>> ar2 = downsample(x,4);
>> length(ar1)
ans =

     1000000
     
>> length(ar2)
ans =

      250000
      
>> ar2 = hilbert(ar2);
      
      % Now how to plot ar1 and transformed ar2 on same time x axis?

Two ideas I had were: repelem, which would just make any given array2 value repeat x number of times, or upsample(), but this introduces 0's into the array. I'd rather have a smooth line if possible. Thank you!

Answer 1

[DGM]

There are probably tools to do this in SPT, but I don't see why interpolation won't work.

Fs = 2000
t = 500
ar1 = 1:1:Fs*t;
ar2 = downsample(ar1,4);

nc = numel(ar2);
nf = numel(ar1);
ar2f = interp1(linspace(1,nf-3,nc),ar2,1:nf);
numel(ar2f) % now ar2f has the same length as ar1