how can i convert 8760 hourly data to ten minute data ?

how can i convert 8760 hourly data to ten minute data ?

Answer 1

[dpb]

Find the underlying database and resample it, perhaps?

If you don't have finer granularity than hourly samples, there's nothing to be done other than interpolate which is in the end nothing more than what you have interpolated to a different resolution; how closely that would follow the actual data that would have been recorded is dependent entirely upon how regular that data might actually be and the choice of interpolating scheme.

[ABDULLA RAHIL]

so the error will increase , could u tell me how can i do interpretation to get 10 minutes

[dpb]

'Pends on how you've got the data to some extent but the basic idea is simply

L=length(Y);  % current dataset length
newT=linspace(T(1),T(end),6*L);  % new time points 6/hr
newY=interp1(T,Y,newT);          % interpolated values

See

doc interp1  % for details

NB: In releases from R2012 and earlier, interp1 worked as the above if Y were an array and x a vector interpolating over the columns as observations. For some reason that is totally unfathomable to me, TMW broke this obvious and extremely useful functionality at some point after that (R2012b is latest release I have so not sure when it no longer will work at all despite warnings).

If you have later version, you must use a looping structure and do the interpolation column-by-column instead... :(

[ABDULLA RAHIL]

sorry , what is the value of T in my case

[dpb]

It's a placeholder for your time vector--you didn't provide any details.