retrieve data specific date

hello, i'm very new in matlab.

I have 2000-2003 hourly nc files.

I have 3 questions.

1. I want to retrieve data per 10 days each month

d1= 1,2,..,10

d2=11,12,..,20

d3=21,22,..31

2. like the first question but I want data in 1 year, I mean

data=[d1 jan, d1 feb,..,d1 dec]

3. I want to retrieve data only 1 time. For example, March 4, 2002 at 15.00

here I include a bit of my nc file.

thanks for the help

Answer 1

[KSSV]

You need to load the nc data into MATLAB. For this you need to use ncdisp and ncread.

You read the dates and convert them into the class datetime. For this you need to read about datetime, datenum and datestr. Once you have dates and data in hand, you can play with the indices and get the data you want.

I will make the datetime data and answer your questions.

Answer 1:

thedates = [datetime(2022,1,1):datetime(2022,12,31)]' ;  % this is the datetime data, you need to get from nc files
class(thedates)

ans = 'datetime'

% pick each 10 days data for Jan 
idx = thedates.Month==1 ;
t = thedates(idx) ; % dates for Jan month alone 
idx1 = t.Day<=10 ;
t(idx1)

ans = 10×1 datetime array
   01-Jan-2022
   02-Jan-2022
   03-Jan-2022
   04-Jan-2022
   05-Jan-2022
   06-Jan-2022
   07-Jan-2022
   08-Jan-2022
   09-Jan-2022
   10-Jan-2022

idx2 = t.Day > 10 & t.Day <=20  ;
t(idx2)

ans = 10×1 datetime array
   11-Jan-2022
   12-Jan-2022
   13-Jan-2022
   14-Jan-2022
   15-Jan-2022
   16-Jan-2022
   17-Jan-2022
   18-Jan-2022
   19-Jan-2022
   20-Jan-2022

Answer 2:

idx3 = thedates.Day == 1 ;
thedates(idx3)

ans = 12×1 datetime array
   01-Jan-2022
   01-Feb-2022
   01-Mar-2022
   01-Apr-2022
   01-May-2022
   01-Jun-2022
   01-Jul-2022
   01-Aug-2022
   01-Sep-2022
   01-Oct-2022
   01-Nov-2022
   01-Dec-2022

Answer 3:

idx4 = thedates.Month == 3 & thedates.Day == 4 ;
thedates(idx4)

ans = datetime
   04-Mar-2022

Like above, you can get the indices. The same indices should be used to extract the data.

[Den]

thank you so much, such a big help

[Peter Perkins]

Den, KSSV is correct in recommending datetime, but DON'T read about datenum and datestr. They are no longer recommended.

Answer 2

[Peter Perkins]

In addition to what KSSV shows, since you say "retrieve data", I recommend that you use a timetable, and things like timerange,or subscripting with datetimes to selec t subsets of your data.

It's possible that some of your uses would benefit from having a categorical variable, not a datetime. For example

>> t = datetime(2022,1,1:365);
>> dayBin = discretize(day(t),[1 11 21 31],"categorical");
>> categories(dayBin)
ans =
  3×1 cell array
    {'[1, 11)' }
    {'[11, 21)'}
    {'[21, 31]'}