How to create a only a time vector starting 13:15:00 to 20:59:00 with 1 minute interval? R2016a

I already have a DATE series and i need to create a time series wiith equally spaced 1 minute interval from 13:15:00 to 20:59:00 and attach it with DATE series. How to Create such time series and then attach them with DATE data.

Answer 1

[Chad Greene]

This is a datetime for today, starting at 13:15 and going to 20:59, in 1 minute [1/(24*60)] intervals.

t = datetime(2018,12,11,13,15,0):1/(24*60):datetime(2018,12,11,20,59,0)

And to check:

>> t(2)-t(1)
ans = 
  duration
   00:01:00

[Steven Lord]

I recommend using minutes(1) instead of using the "magic number" 1/(24*60).

[Chad Greene]

Nice!

[Steven Lord]

Actually, rereading the poster's question, I'd start off with a duration array.

>> startDur = duration(13, 15, 0);
>> endDur = duration(20, 59, 0);
>> t2 = startDur:minutes(1):endDur;

Then you can add that to a datetime representing the start of any day, like today.

>> startOfToday = datetime('today');
>> t3 = t2 + startOfToday;
>> t3(1:10).'

Or you could use January 1st of this year.

>> startOfYear = dateshift(startOfToday, 'start', 'year');
>> t4 = t2 + startOfYear;
>> t4(1:10).'

[NS]

Thank you sir , I am able to create a duration from 13:15 to 20:59 but I need to stack them to a list of DATES of format 'MM/dd/yyyy'. How to do this as STACK doesnt seem to work for datetime.

Actually my datetime series has some missing minutes which I need to interpolate. So for that first I need a complete time vector from 13:15 to 20:59 and then combine them with my list of DATES as some dates like weekends and holidays are deleted so t4=t2+ startofyear wont work in my case as my dates are not all consecutive.

How to stack this time vector with my DATE vector?

[NS]

Any help on this?

[Steven Lord]

What exactly do you mean by "stack this to a list of DATES of format 'MM/dd/yyyy'"? Show us what it means to "stack" this duration vector:

dur = duration(13, 0:15:60, 0)

with this vector of datetime objects:

dat = datetime('today') + days(0:5)

What exactly do you expect / need / want the result to be?

[NS]

I have a DATETIME series begining 01/01/2010 13:15:00 to 31/12/2012 20:59:00 and I have cleaned this DATETIME for weekends, holidays, trading hours because of which I have missing minutes on some days in interval 13:15 to 20:59 , so i need to interpolate all minutes in this interval for the given cleaned DATES (which might not be consecutive due to missing days).

How to interpolate those minutes for those given DATES ?

The output I need is all 465(8*60-15) minutes for all DATES in my cleaned sample.

[Steven Lord]

See if the "Interpolation of Dates and Times" example on the documentation page for the interp1 function does what you want.

Alternately if your data is stored in a timetable consider calling retime on it as shown in the examples on this documentation page.

[NS]

Sir I am working with 2016a version of Matlab and hence i dont have timetable or its related functionality.

I am trying with interp1 but still not getting desired result .

Can you please help with an example?

[Steven Lord]

Show us the interp1 call you tried (along with the code to create a small sample of data that we can use to run that interp1 call) and we may be able to help you figure out how to get it to do what you want.

[NS]

Sir I have attached a sample data whherein my first column is DATETIME which are cleaned for non trading hours, holidays and weekends, outliers and second column is PRICE.

My purpose is to get all MINUTES between 13:15 to 20:59 on all DATES in cleaned sample DATES and interpolate the PRICES . I have tried -

interp1(DATETIME(1),PRICE,DATETIME(end))

But I am not getting desired result . Kindly help . I have attached sample data in excel.

[NS]

Thanks sir , I figured it out.