datestr problem with minutes

Hello everyone,

I'm encountering a problem with date conversion. I have a dataset column with dates in '1/5/2017 12:10:00AM' form and I want to convert it into '2017-05-01 00:10:00' format. I used the datestr function but the minutes are incorrect. All rows return with 00 minutes.

How can I fix this?

Thank you in advance, Vanessa

[Stephen23]

@Vanessa: please show us the code you have used and the data that causes this behavior.

[Vanessa]

kwhm.date=datestr(kwhm.date,'yyyy-mm-dd HH:MM:SS');

date '1/5/2017 12:10:00AM' '1/5/2017 12:15:00AM' '1/5/2017 12:20:00AM' etc

[Andrei Bobrov]

Dear Vanessa! please accept the answers that solved your problems.

Answer 1

[Walter Roberson]

t = datenum('1/5/2017 12:10:00AM' );
datestr(t, 'YYYY-mm-DD HH:MM:ss')

Answer 2

[Andrei Bobrov]

a = datetime('1/5/2017 12:10:00AM','I','dd/MM/yyyy hh:mm:ssa','F','yyyy-MM-dd HH:mm:ss')

[Vanessa]

Thanks this worked but I have some rows which are '1/5/2017' without the time and the result of the above function is NaT. How can overcome this? thanks

[Walter Roberson]

You cannot. When you use datetime() all of the entries need to be in the same 'InputFormat' (that here was abbreviated as 'I'). If you know you have entries that are in a different format, you need to detect and process them.

For example:

a = datetime(StringArrayOfTimes, 'I', 'dd/MM/yyyy hh:mm:ssa','F','yyyy-MM-dd HH:mm:ss');
failed_on = isnat(a);
a(failed_on) = datetime(StringArrayOfTimes(failed_on), 'I', 'dd/MM/yyyy', 'F', 'yyyy-MM-dd HH:mm:ss');

The above does a patch-up pass, where it tries to re-interpret the failed times with the abbreviated format.

[Vanessa]

Thank you very much!!!

[Andrei Bobrov]

a = {'1/5/2017';
    '1/5/2017 12:10:00AM';
    '1/5/2017 12:15:00AM';
    '1/5/2017 12:20:00AM';
    '2/5/2017'};

t = ~cellfun(@isempty,regexp(a,'\d{4}$'));
a(t) = cellfun(@(x)[x,' 00:00:00AM'],a(t),'un',0);
out = datetime(a,'I','dd/MM/yyyy hh:mm:ssa','F','yyyy-MM-dd HH:mm:ss')