This works:
[d,s,r] = xlsread('Damith test.xls');
DN = datenum([d(:,1:4) repmat([0 0], size(d,1), 1)]); % Convert All To ‘datenum’
D0 = floor(DN(1)); % Start Date & Time
DE = ceil(DN(end)); % End Date & Time
DT = 1/24; % Days/Hour
DV = D0 + cumsum([0; ones(round(24*(DE-D0)),1)*DT]); % All Hours
DV = DV(1:3:end-1); % q3 Hours
DC = unique([DN; DV]); % Combine Date Vectors
DataIntrp = interp1(DN, d(:,5:6), DC); % Interpolate
DateMtx = datevec(DC); % Date Vectors
DataResult = [DateMtx(:,1:4), DataIntrp];
DataResult = DataResult(~isnan(DataResult(:,5)),:); % Output Eliminating Extrapolations
CheckResult = DataResult([1:10 end-9:end],:) % Check Result (Temporary)
The ‘CheckResult’ line is there to look at the first and last 10 entries. It doesn’t match your ‘B’ matrix, but it does match your description (and your question in my not yet Accepted Answer to your previous Question). It could likely be made more efficient, but if you only need to run it once for each data set, this is likely sufficient.
