Dealing with leap years, creating arrays of yearly data

I have a time series in a Nx1 array where N are the number of days (different for each dataset, but lets take days from 1951 till 2007). Leap years are included (1952, 1956, ...). I am trying to convert the array into 366x57 (57 being the number of years between 1951-2007), by adding Nan at the Feb 29th (60th) position for non-leap years. I am sure there is a simple solution which I am having a hard time devising. Is there a way to accomplish this without loops.

Answer 1

[the cyclist]

Here is an example using two years, one leap and one not. Should be easy for you to see how to generalize.

% Time series with one leap year and one not
T = rand(365+366,1);

% The years
Y = 1951:1952;
numberYears = numel(Y);

% Slick way to identify the leap years
isLeapYear = datenum(Y,2,29)~=datenum(Y,3,1);

T_new = nan(366,numberYears);

for ny = 1:numberYears
    numberDaysThisYear = 365+isLeapYear(ny);
    T_new(1:numberDaysThisYear,ny) = T(1:numberDaysThisYear);
    T(1:numberDaysThisYear) = []; % Deletes T as you go. Could do this differently
end

[SMA]

Great! That is slick.

Can I have the NaN added at 60th entry instead of end of the matrix.

[the cyclist]

I think I got the alignment right here, but you should double-check:

% Time series with one leap year and one not
T = rand(365+366,1);

% The years
Y = 1951:1952;

% Slick way to identify the leap years
isLeapYear = datenum(Y,2,29)~=datenum(Y,3,1);

numberYears = numel(Y);

T_new = nan(366,numberYears);

for ny = 1:numberYears
    numberDaysThisYear = 365+isLeapYear(ny);
    if isLeapYear(ny)
        T_new(1:366,ny) = T(1:366);
    else
        T_new(1:59,  ny) = T(1:59);
        T_new(61:366,ny) = T(60:365);
    end
    T(1:numberDaysThisYear) = []; % Deletes T as you go. Could do this differently.
end

[dpb]

"% Slick way to identify the leap years"

My favorite is one of my standard utilities...

function is=isleapyr(yr)
%  returns T for input year being a leapyear

is=eomday(yr,2)==29;