I have issues using datenum function because of a different date format on excel, how can I modify my dates in matlab and then run the function?n

EX is the data table (16888x3) and the date time format in the first column of the table is "HH:MM:SS, dd mm yyyy"

X=EX.DateTime

Unable to resolve the name 'EX.DateTime'.

Y1=EX.PM25
Y2=EX.Humidity

DT= datenum(X,"HH:MM:SS, dd mm yyyy")

[Tanishq]

i have to do time series analysis of a large amount of data, please help asap!

[Chunru]

Attach some data sample.

[Tanishq]

Thanks Mr. chunru for replying so fast but Actually to share the data I have to take permission from someone but I'll be right back.

Answer 1

[Steven Lord]

Don't use serial date numbers and datenum. Use datetime instead.

s = "12:34:56, 30 08 2022"

s = "12:34:56, 30 08 2022"

fmt = "HH:mm:SS, dd MM yyyy";
dt = datetime(s, InputFormat=fmt) % Use the default Format

dt = datetime
   30-Aug-2022 12:34:00

dt2 = dt; dt2.Format = fmt % or you can use the original format

dt2 = datetime
   12:34:56, 30 08 2022

[Tanishq]

Thanks, I'll try using this!

Answer 2

[dpb]

Don't use datenum, use datetime instead and readtable,

You can either use an import options object from detectImportOptions in which you set the format for the date format or let it read it as cellstr or string and convert.

Presuming you've done the latter to get the posted code -- it would have been more helpful to have seen that as well as the short example section of the actual input file...

EX.DateTime=datetime(EX.DateTime,'InputFormat','HH:mm:ss, yyyy-MM-dd');
EX.Time=EX.DateTime-EX.DateTime(1);             % convert to duration for use in analysis

You may also find it convenient to use seconds(EX.Time) as the analysis time variable for timeseries analyses instead of the duration class, depending upon just what functions you're planning on using; not all of the Signal Processing TB has been made datetime/duration aware.

[Tanishq]

Thanks for the valuable input Mr. dpb