interp1 for matrix

I have two velocity matrices that I want to interpolate onto a new time vector. The matrices are both 8011x60 doubles where the rows are different times and the columns are depths. Can I use interp 1 to interpolate each row onto the new time?

This is how I have it set up now (how I would use interp1 if U and V were vectors):

I took a screenshot of my workspace too if that helps.

U_new = interp1(Date,U,M1time);
V_new = interp1(Date,V,M1time);
Temp_new = interp1(Date,Temp,M1time); 

[dpb]

Did you not try it? There's an example in the documentation precisely the usage to interpolate arrays by column to a common set of query points...

[Heidi Hirsh]

Hi! Thank you for pointing out the size discrepancies. I realized I had somehow reloaded an old velocity file (that was two timestamps short). Once I fixed that, interp1 worked fine the way I had written it. I apologize for appearing so idiotic. Thank you for your help :)

Answer 1

[darova]

You have to make the same number of rows for Date and U

U_new = interp1(Date,U,M1time);
    % size(Date): 8013x1
    % size(U):    8011x60    

Didn't you get an error about size match?

[Heidi Hirsh]

I fixed the size issue and it ran fine. I somehow reloaded the wrong data file (2 velocity vectors shorter than time).

Answer 2

[Walter Roberson]

No you should use interp2 for this. The value to look up depends on two factors, depth and time, but interp1 would only take one factor into account.

[dpb]

Hmmmm, Walter. I don't see what you're seeing--just that wants to interpolate onto a new time vector the two arrays. Altho I see the reference to the depth being what the columns of the array refer to I didn't see and his example code doesn't seem to want to do more than just those onto the Mitime vector.

Guess we'll find out if OP comes back... :)

[Heidi Hirsh]

Hi guys! Thank you for the help! I think in this case interp1 did the trick since I was only trying to interpolate over time. But I have never used interp2 before so I really appreciate that advice and I will try it on another data set I have in mind where I do need to interpolate over depth.