Adding start date to timestamps

Question

Oliver Higbee 2019-3-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/448398-adding-start-date-to-timestamps

评论： Oliver Higbee 2019-4-25

I'm importing data from a .csv file containing temperatures logged every 30 seconds for 4 days. There is a column containg timestamps but with no date information.

I would like to import this in matlab and add dates to the timestamps, I'm using 'datetime' to get it into the format I would like using;

conductor_temps=xlsread('1March19to5March19_data_set2.csv');
TimeStamps = datetime(conductor_temps(:,2), 'ConvertFrom', 'excel');

This assigns 31-Dec-1899 to each timestamp due to the lack of date information. I can use;

TimeStamps.Day = 1;
TimeStamps.Month = 3;
TimeStamps.Year = 2019;

And manually assign. Is there a more elegant way to do this. For example assign a start date to the dataset and have it move to the following day each time the timestamps reaches midnight?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Geoff Hayes 2019-3-5

Oliver - you could try adding a reference datetime to your array. See date and time arithmetic for details.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Peter Perkins 2019-3-11

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/448398-adding-start-date-to-timestamps#answer_364839

编辑：Peter Perkins 2019-3-12

在 MATLAB Online 中打开

Oliver, it's not 100% clear what you have, but I think this will do what you want. The main issue is, I guess, that your spreadsheet is formatted kind of funny, with no date info at all. I'm showing this done in raw workspace variables, but I actually recommend that you use readtable and work in the table you get back.

>> timestamps = repmat([0 43200]'/86400,3,1) % what I think you have
timestamps =
         0
    0.5000
         0
    0.5000
         0
    0.5000
>> t = datetime(timestamps,'ConvertFrom','Excel')
t = 
  6×1 datetime array
   31-Dec-1899 00:00:00
   31-Dec-1899 12:00:00
   31-Dec-1899 00:00:00
   31-Dec-1899 12:00:00
   31-Dec-1899 00:00:00
   31-Dec-1899 12:00:00
>> newday = [true; diff(t)<0]
newday =
  6×1 logical array
   1
   0
   1
   0
   1
   0
>> daynum = cumsum(newday)
daynum =
     1
     1
     2
     2
     3
     3
>> date = datetime(2019,3,1) + caldays(daynum-1)
date = 
  6×1 datetime array
   01-Mar-2019
   01-Mar-2019
   02-Mar-2019
   02-Mar-2019
   03-Mar-2019
   03-Mar-2019
>> time = t - datetime(1899,12,31) % Excel serial day number epoch
time = 
  6×1 duration array
   00:00:00
   12:00:00
   00:00:00
   12:00:00
   00:00:00
   12:00:00
>> dt = date + time
dt = 
  6×1 datetime array
   01-Mar-2019 00:00:00
   01-Mar-2019 12:00:00
   02-Mar-2019 00:00:00
   02-Mar-2019 12:00:00
   03-Mar-2019 00:00:00
   03-Mar-2019 12:00:00