datestr datenum problem with year conversion.

Hello I have the following issue,

I am pulling out data from a Quandl, I get the following the set :

The syntax to pull out is the following :

 Quandl.get('USTREASURY/YIELD', 'trim_start', '1990-01-02', 'trim_end', '2014-09-12', 'authcode', 'jtwf2XzCdzxhAKrXxNkm')

Time Series Collection Object: USTREASURY/YIELD

Time vector characteristics

      Start date            02-Jan-1990 00:00:00
      End date              12-Sep-2014 00:00:00

Member Time Series Objects:

      x1Mo
      x3Mo
      x6Mo
      x1Yr
      x2Yr
      x3Yr
      x5Yr
      x7Yr
      x10Yr
      x20Yr
      x30Yr

Say this set is called YC.

By typing YC.Time or YC.x1Mo.Time, I get the datenum but it allocated the start date which is 02-Jan-1990 00:00:00 as 0 then increment till the end date which is 9019.( 9 sept 2014)

See a sample here

0 1 2 3 6 7 8 9 10 14 15 16 17

When i make a datestr of this matrix, I get a straight integer increment for year 1990, to now, that is 1990 is taken as 0 and ends as 0024, which stands for 2014.

See Sample

00/01/0000 00:00:00 01/01/0000 00:00:00 02/01/0000 00:00:00 03/01/0000 00:00:00 06/01/0000 00:00:00 07/01/0000 00:00:00 08/01/0000 00:00:00 09/01/0000 00:00:00 10/01/0000 00:00:00 14/01/0000 00:00:00 15/01/0000 00:00:00 16/01/0000 00:00:00 17/01/0000 00:00:00

I have tried also things like this datestr(YC.Time,'dd/mm/yyyy HH:MM:SS') but does not work.

I have the impression, the year is non existent in the time vector and it has only month and date and allocates an incremental year number till now.

Do you know how I can make this time vector starts from 1990 and continue till now.

I can code something to make this work in a loop, but I fear there is a much easier way to do that using may be pivot year, which I cannot seem to find.

Thanks you very much.

D.

Answer 1

[Chad Greene]

Is this what you're looking for?

months = [0 1 2 3 6 7 8 9 10 14 15 16 17];
datestr(datenum(1990,1+months,2))

Answer 2

[dpb]

Time series objects seemingly use the differential time internally; use the getabstime method to return the actual time vector it seems. I've not used it so this is just reading the doc's...

http://www.mathworks.com/help/matlab/ref/timeseries.getabstime.html