Efficiently find most recent value using TIME column in two tables

Hi everyone,

I have two large tables (> 10 million rows), one containing measurement data ("data")

• each data point with a time stamp, approximately 2 Hz but significant jitter

one containing reference data ("ref")

• also with one time stamp per data point, approximately 1.5 Hz but significant jitter as well

I'd like to add a column to "data" named data.ref which always contains the most recent reference point.

The following works, but is incredibly inefficient:

for k=1:numel(data.TIME)
    i_mostRecent = find(ref.TIME <= data.TIME(k), 1, 'last');
    data.ref(k) = ref.ref(i_mostRecent);
end

Any tips on how to get rid of the for-loop and speed things up? Thanks a lot in advance!

Answer 1

[Hans Meier]

Ok, here's one solution:

data.ref = interp1(ref.TIME, ref.ref data.TIME, 'previous');