Seasonal Variability of time series

I would like to find out the seasonal variability of the data. My data is form 1883-2018 temperature. Each month has several data. I tried to calculate the average of each month and do the seasonal variation. But I got NaN for sst. Here is my code.

MIDAS00525temp = load('../TimeseriesProject/MIDAS_srcid00525_daily_temperature.txt');

tempmin = MIDAS00525temp (:,3);

tempmin(tempmin<0) = NaN; MATLABDate_temp = MIDAS00525temp(:,2);

tempfilled = inpaint_nans(tempmin, 3);

[Y,M] = datevec(MIDAS00525temp(:,2));

[a,~,c] = unique([Y,M],'rows'); yearaveragemin = [a, accumarray(c,MIDAS00525temp(:,3),[],@(x)mean(x))];

datayearaveragemin = (yearaveragemin(:,3));

MATLABDate_temp = datenum(yearaveragemin(:,1), yearaveragemin(:,2), 0);

% Make a moving average

temp_medfiltmin = medfilt1(datayearaveragemin, 36, 'omitnan', 'truncate'); xt = datayearaveragemin - temp_medfiltmin; T = length(MATLABDate_temp);

% Create seasonal indices

s = 12; % number of timesteps per year sidx = cell(s,1); for i = 1:s sidx{i,1} = i:s:T; end % Apply a stable seasonal filter.

sst = cellfun(@(x) mean(xt(x)),sidx);