How to populate complete time series with incomplete data

4 次查看（过去 30 天）
Colin Edgar
Colin Edgar 2016-1-21
回答： Colin Edgar 2016-1-21
I have time series data with skipped periods that I wish to place onto a complete time stamp array. Rather than mm/dd format I have 'year' 'day of year (DOY)' and 'HHMM' 'Var1' like so
data=
2015 59 30 -31.42
2015 59 100 -31.21
2015 59 130 -31.06
2015 59 700 -30.9
2015 59 730 -30.38
2015 59 800 -30.07
2015 59 830 -29.83
2015 59 900 -29.69
2015 59 930 -29.53
2015 59 1000 -29.27
2015 59 1500 -26.11
2015 59 1530 -26.09
By comparing 'data' with a complete time series array 'full', I can get an 'index' value for missing periods, but I can't figure out how to bring the data (Var 1 etc) with it.
index = ismember( full(:, 3), data(:,3));
Only how to overlay the full set with NaN values for missing indices
fullindex = [index index index];
full(~fullindex)=nan;
Similar to this post http://www.mathworks.com/matlabcentral/answers/175089#answer_166928 but I can't use the 'datenum()' approach since I have DOY and different time stamps.

请先登录，再回答此问题。

采纳的回答

Chad Greene
Chad Greene 2016-1-21
You can get a vector in datenum format by
t = datenum(YEAR,1,DOY);
and if you'd like you can get t into year, month, day by
[year,month,day] = datevec(t);
  1 个评论
Colin Edgar
Colin Edgar 2016-1-21
You have no idea how helpful this is, thanks! I didn't understand datenum() at all apparently.
I see that if the HHMM is split up that I can have it all:
t = datenum(YEAR,1,DOY,HOUR,MINUTE,1);

请先登录，再进行评论。

更多回答（1 个）

Colin Edgar
Colin Edgar 2016-1-21
Complete solution to problem in OP:
%Change input 'data' and 'full' to separate HH and MM into separate columns
%So Year, DOY, Hour, Min, Var1
%Create Var1 as NaN in 'full'
%get time vectors for 'data' and 'full'
t = datenum(Year, 1, DOY, Hour, Minute, 1);
ft = datenum(Year, 1, DOY, Hour, Minute, 1);
datadates = round(datevec(t));
fulldates = round(datevec(ft));
[~, index] = ismember(datadates, fulldates, 'rows');
full(index,:) = data;
Thanks to Guillaume for solving the rest of this here: http://www.mathworks.com/matlabcentral/answers/184017#answer_171949

类别

在 Help Center 和 File Exchange 中查找有关 Dates and Time 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by