Try something like this —
Date = datetime(2020,01,01)+caldays(0:3.9*365.25).';
Temp = 80*rand(size(Date))-40;
Humd = randi([1 100],size(Date));
Wx = timetable(Date,Temp,Humd) % Original 'timetable'
WxYr = retime(Wx, 'yearly','max') % Yearly Maxima
Yrs = year(WxYr.Date);
for k = 1:numel(Yrs)
Lv1 = year(Wx.Date) == Yrs(k); % Logical Vector Selecting Days For Selected Year
WxCurYr = Wx(Lv1,:); % Current Year 'timetable'
Lv2 = ismember(WxCurYr.Temp, WxYr(k,:).Temp); % Select Variable To Compare (Here: Temperature) & Return Logical Vector Of Matches
ResultTT{k,:} = WxCurYr(Lv2,:); % Save 'timetable' Arrays
Result(k,:) = timetable2table(WxCurYr(Lv2,:)); % Save Rows In The 'Result' Table
end
Result % Table Of Days Matching Yearly Maximum Temperature
If there is more than one matching value, this becomes slightly more complicated. In that instance, it will be necessary to either return all of the matching days (this usually requires that ‘result’ becomes a cell array, causing further complications later in the code), or use find with ‘Lv2’ and use an element of that vector to return only one of them (the approach I have characteristically taken in these situations, simply because it’s easier).
.
