time date to datenum and datenum to datetime

I have time in 'yyyy-MM-dd''T''HH:mm:ss.SSS''Z' formate (e.g., [2018-08-01T20:05:00.000Z; 2018-08-01T20:10:00.000Z; 2018-08-01T20:35:00.000Z; 2018-08-01T25:05:00.000Z] . How to make time string and datetime to datenum and datenum to datetime?

Thanks in advance.

(I am new in matlab.)

[Stephen23]

What is the particular need to use serial date numbers?

In general you should avoid deprecated serial date numbers and use DATETIME/DURATION-based code.

[Ismita]

But I need it serially here

[Jan]

25:05 is a strange time.

Please post some valid Matlab code, which creates exactly the input you have. "[2018-08-01T20:05:00.000Z; 2018-08-01T20:10:00.000Z; 2018-08-01T20:35:00.000Z; 2018-08-01T25:05:00.000Z]" is no valid code, so the readers must invent your data at first and this is fragile.

[Stephen23]

"But I need it serially here"

What is the exact operation that requires a serial date number?

Both DATETIME and DURATION objects support many many methods and functions, if you actually explained how you are going to process your data then we could help you use a better approach.

[Ismita]

Thank you. You can check the data attached in the comment threading.

Answer 1

[Bruno Luong]

I make a round conversion so you can convert any format to any other by using the appropriate substeps

dt=datetime(2022,04,28,13,09,00,'Format','yyyy-MM-dd''T''HH:mm:ss.SSS''Z''')

dt = datetime
   2022-04-28T13:09:00.000Z

dn=datenum(dt)

dn = 7.3864e+05

ds = datestr(dn)

ds = '28-Apr-2022 13:09:00'

dt = datetime(ds,'InputFormat','dd-MMM-yyyy HH:mm:SS','Format','yyyy-MM-dd''T''HH:mm:ss.SSS''Z''')

dt = datetime
   2022-04-28T13:09:00.000Z

[Ismita]

Thank you. How can I make such a column like dt? Can I use split/ something to make the column?

[Bruno Luong]

I don't understand. A column of what? from what? What is "split/ something"?

[Jan]

@Mst Ismita Tasnim: I've asked you already, what exactly your input arguments are. Please post, how your data look.

[Ismita]

Thank you so much. sorry for being late. Time is "'yyyy-MM-dd''T''HH:mm:ss.SSS''Z'", e.g.,

2008-01-01T00:00:00.000Z -1.06

2008-01-01T00:05:00.000Z -1

2008-01-01T00:10:00.000Z -1.76

2008-01-01T00:15:00.000Z -2.09

2008-01-01T00:20:00.000Z -1.74

2008-01-01T00:25:00.000Z 999

2008-01-01T00:30:00.000Z -3.64

2008-01-01T00:35:00.000Z -3.65

2008-01-01T00:40:00.000Z 999

2008-01-01T00:45:00.000Z -3.68

2008-01-01T00:50:00.000Z -3.98

2008-01-01T00:55:00.000Z -3.83

2008-01-01T01:00:00.000Z -3.48

2008-01-01T01:05:00.000Z -3.45

2008-01-01T01:10:00.000Z -3.24

2008-01-01T01:15:00.000Z -2.4

[Bruno Luong]

@Mst Ismita Tasnim Sigh..., What are these data? Do you copy from MATLAB command line? You excel file? Your notebook? Do you hire someone to type in the computer for you?

If it's come from MATLAB data, please explain us the type/class/size, or save it as mat file and attach here.

What do you want to do with it?

PLEASE EXPLAIN!

[Bruno Luong]

In the mean time you explain, here is an exemple with vector of datetime

dt=datetime(2022,04,28,12+randi(12,5,1),09,00,'Format','yyyy-MM-dd''T''HH:mm:ss.SSS''Z''')

dt = 5×1 datetime array
   2022-04-28T17:09:00.000Z
   2022-04-28T23:09:00.000Z
   2022-04-28T20:09:00.000Z
   2022-04-28T17:09:00.000Z
   2022-04-29T00:09:00.000Z

dn=datenum(dt)

dn = 5×1

1.0e+05 * 7.3864 7.3864 7.3864 7.3864 7.3864

ds = datestr(dn)

ds = 5×20 char array
    '28-Apr-2022 17:09:00'
    '28-Apr-2022 23:09:00'
    '28-Apr-2022 20:09:00'
    '28-Apr-2022 17:09:00'
    '29-Apr-2022 00:09:00'

dt = datetime(ds,'InputFormat','dd-MMM-yyyy HH:mm:SS','Format','yyyy-MM-dd''T''HH:mm:ss.SSS''Z''')

dt = 5×1 datetime array
   2022-04-28T17:09:00.000Z
   2022-04-28T23:09:00.000Z
   2022-04-28T20:09:00.000Z
   2022-04-28T17:09:00.000Z
   2022-04-29T00:09:00.000Z

[Ismita]

I have attached the txt file that I build from csv data set.

[Bruno Luong]

Thanks goodness finally a data file.

Import and convert to datetime as below, do the rest as with my earlier answer

s=readtable('File.txt')

s = 10×10 table
                Var1                Var2     Var3     Var4      Var5     Var6    Var7     Var8    Var9     Var10
    ____________________________    _____    _____    _____    ______    ____    _____    ____    _____    _____

    {'2008-01-01T00:00:00.000Z'}    -1.06      4.3      1.2    -355.9      -9    -16.6    6.58    66335       0 
    {'2008-01-01T00:05:00.000Z'}       -1     3.89     1.26    -353.4    -6.7      -18     6.6    71037     300 
    {'2008-01-01T00:10:00.000Z'}    -1.76     3.22     -0.7      -350     0.6    -11.9    6.59    74186     600 
    {'2008-01-01T00:15:00.000Z'}    -2.09     3.29    -0.49    -349.5      -2     -4.4    6.62    88787     900 
    {'2008-01-01T00:20:00.000Z'}    -1.74      2.3    -1.49    -351.4    -2.6     -3.7    6.65    82061    1200 
    {'2008-01-01T00:25:00.000Z'}    -3.27     0.41    -2.61    -346.2     2.8     -0.3    6.47    74540    1500 
    {'2008-01-01T00:30:00.000Z'}    -3.64    -0.44    -2.43    -343.7     8.4      3.5    6.09    82033    1800 
    {'2008-01-01T00:35:00.000Z'}    -3.65     0.27    -1.81    -343.2    12.4        0    6.35    82179    2100 
    {'2008-01-01T00:40:00.000Z'}     -3.4     1.72    -1.67    -341.8     5.7      5.2    6.15    77514    2400 
    {'2008-01-01T00:45:00.000Z'}    -3.68     0.59    -1.39    -343.1     0.5     -4.7    6.27    80708    2700 

ts=table2cell(s(:,1))

ts = 10×1 cell array
    {'2008-01-01T00:00:00.000Z'}
    {'2008-01-01T00:05:00.000Z'}
    {'2008-01-01T00:10:00.000Z'}
    {'2008-01-01T00:15:00.000Z'}
    {'2008-01-01T00:20:00.000Z'}
    {'2008-01-01T00:25:00.000Z'}
    {'2008-01-01T00:30:00.000Z'}
    {'2008-01-01T00:35:00.000Z'}
    {'2008-01-01T00:40:00.000Z'}
    {'2008-01-01T00:45:00.000Z'}

dtformat = 'yyyy-MM-dd''T''HH:mm:ss.SSS''Z';
dt = datetime(ts,'Format',dtformat,'InputFormat',dtformat)

dt = 10×1 datetime array
   2008-01-01T00:00:00.000Z
   2008-01-01T00:05:00.000Z
   2008-01-01T00:10:00.000Z
   2008-01-01T00:15:00.000Z
   2008-01-01T00:20:00.000Z
   2008-01-01T00:25:00.000Z
   2008-01-01T00:30:00.000Z
   2008-01-01T00:35:00.000Z
   2008-01-01T00:40:00.000Z
   2008-01-01T00:45:00.000Z

[Ismita]

Thank you so much. If I use 's=readtable('File.txt')', do Var2, Var3, Var4,.... behave like text/string? I mean do I need to use anything like Var2 = str2double(s(:,2:9)) or 'table2array(s(:,2:9))' ?

[Bruno Luong]

They are table, if you aceess them by column name, it returns the standard column-vector of doubles

s.Var2 + s.Var3

[Ismita]

thanks a lot.

Answer 2

[Steven Lord]

I'm going to assume that you've imported this data into MATLAB as a string array or a cellstr with two columns. Since I don't have the file or whatever other source this data comes from, I'll hard code it so my example can operate on it.

S = ["2008-01-01T00:00:00.000Z",	"-1.06";
     "2008-01-01T00:05:00.000Z",	"-1";	
     "2008-01-01T00:10:00.000Z",	"-1.76";	
     "2008-01-01T00:15:00.000Z",	"-2.09"];

Let's create a datetime array using the first column of S.

fmt = 'yyyy-MM-dd''T''HH:mm:ss.SSS''Z''';
dt = datetime(S(:, 1), 'InputFormat', fmt, 'Format', fmt)

dt = 4×1 datetime array
   2008-01-01T00:00:00.000Z
   2008-01-01T00:05:00.000Z
   2008-01-01T00:10:00.000Z
   2008-01-01T00:15:00.000Z

Now to create a double array using the second column of S.

n = double(S(:, 2))

n = 4×1

-1.0600 -1.0000 -1.7600 -2.0900

Finally, storing this date and time data along with the numeric data in a timetable array would let us perform future operations on the data easily.

T = timetable(dt, n)

T = 4×1 timetable
               dt                 n  
    ________________________    _____

    2008-01-01T00:00:00.000Z    -1.06
    2008-01-01T00:05:00.000Z       -1
    2008-01-01T00:10:00.000Z    -1.76
    2008-01-01T00:15:00.000Z    -2.09

[Ismita]

Thank you so much. Could you please inform me the type of 'T=timetable(dt,n)' data is text/numeric?

Should I use 'double(n)' / 'table2array(n)' for other variables like 'n' here?

Answer 3

[Corey Silva]

While serial date numbers (datenum) can represent dates and times, it is recommended that you use datetime values to represent points in time, and durationor calendarDuration values to represent elapsed times.