I have a matrix with one column in serial datenumber format, how do I select data from a certain time period?

I have a large dataset where the first column is filled with serial datenumbers. I want to use these to select the rows of data for time periods of interest. For example, the first row in this data set (2) corresponds to the 29th May 2019 and runs til the end of July. I want to extract, for example, data from June only. How do I do this? I know I could convert the column to datetime and then select the rows of interest from there, but is there a simpler way?

[Adam Danz]

"I know I could convert the column to datetime and then select the rows of interest from there, but is there a simpler way? "

What's difficult about that? If you're using >1 line of code to convert that column to datetime, show us what you're doing.

Note that you can also select rows of interest from the date-number format, too.

[Louise Wilson]

Thanks Adam, this is what I am doing right now:

listofdates=tawh(:,1);
listofdates=datetime(listofdates, 'ConvertFrom', 'datenum');
%What rows contain data of interest? 

I then look manually at the data to see which row marks the 1st June and which row marks the 30th June.

Then I do this:

[ra,ca]=size(tawh); %ra is number of rows, ca is number of columns
tawh(2:375,:)=[];
A = tawh(1:4319,1:ca); 
listofAdates=A(:,1);
listofAdates=datetime(listofAdates, 'ConvertFrom', 'datenum');

I want to keep the first row which contains frequency info for each column. So, I select the rows I want to remove, then I have to look back at the data again so see which row the 30th June has moved to. Then, I select this. To confirm, I check the corresponding dates again to make sure I haven't made any errors. This seems preeeetty laborious???

[Louise Wilson]

I could just convert the column to datetime I guess, how would I do this? I planned to keep it as is as I'll use it later, but I guess it would be easier to change it back...

[Adam Danz]

If there's no need to be able to read the dates, you can just use indexing as Stephen Cobeldick explained in his answer. If you need to read the dates, you can convert them and using indexing from datetime format as well.

Answer 1

[Stephen23]

Serial date numbers are, as their name implies, just numbers. So it is very easy to use logical comparisons, just like you would for comparing any other numbers.

First we generate some fake data (because you did not provide any):

>> N = 23;
>> D = linspace(datenum(2019,5,29),datenum(2019,7,31),N);
>> M = [D(:),randi(99,N,8)]
M =
   737574.00000       63.00000       70.00000       22.00000       74.00000        1.00000       27.00000       91.00000       91.00000
   737576.86364       74.00000       36.00000        6.00000       57.00000       28.00000       43.00000       41.00000       72.00000
   737579.72727        7.00000       99.00000       22.00000       56.00000       18.00000       95.00000       67.00000       79.00000
   737582.59091       70.00000       68.00000       33.00000       73.00000       93.00000       87.00000       99.00000       44.00000
   737585.45455       78.00000       57.00000       99.00000       21.00000       42.00000       27.00000       13.00000       22.00000
   737588.31818       19.00000       95.00000       79.00000        4.00000       43.00000       67.00000       15.00000       10.00000
   737591.18182       70.00000       33.00000       75.00000       67.00000       97.00000       46.00000       64.00000       89.00000
   737594.04545       42.00000       48.00000       14.00000       18.00000       61.00000       65.00000       26.00000       48.00000
   737596.90909       62.00000        7.00000       68.00000       18.00000        6.00000       76.00000       72.00000       90.00000
   737599.77273       39.00000       77.00000       59.00000       79.00000       83.00000       77.00000       98.00000       48.00000
   737602.63636       90.00000       93.00000       84.00000        4.00000       84.00000       66.00000       42.00000       21.00000
   737605.50000       74.00000       13.00000       66.00000       12.00000        1.00000       74.00000       64.00000       48.00000
   737608.36364       15.00000       20.00000       34.00000       71.00000        4.00000        1.00000       25.00000       54.00000
   737611.22727       45.00000       65.00000       83.00000       26.00000       19.00000       47.00000        5.00000       53.00000
   737614.09091       15.00000       25.00000        9.00000       94.00000       68.00000       45.00000       57.00000       52.00000
   737616.95455       50.00000       84.00000       52.00000       47.00000       52.00000       82.00000       55.00000       16.00000
   737619.81818       85.00000       33.00000       39.00000       97.00000       70.00000       41.00000       65.00000       41.00000
   737622.68182       44.00000       88.00000        7.00000       16.00000        7.00000       47.00000        7.00000        8.00000
   737625.54545       33.00000       33.00000       44.00000       60.00000       79.00000       60.00000       17.00000       58.00000
   737628.40909       21.00000       29.00000       78.00000       63.00000        5.00000       69.00000       49.00000       68.00000
   737631.27273       37.00000       43.00000       53.00000       72.00000       68.00000       33.00000       27.00000       39.00000
   737634.13636       58.00000       56.00000       96.00000       27.00000       61.00000       89.00000       33.00000       56.00000
   737637.00000       85.00000       58.00000       32.00000       22.00000       99.00000       91.00000       16.00000       80.00000

Then we can use basic MATLAB indexing to select only June's data:

>> idb = M(:,1) >= datenum(2019,6,1); % start of June
>> ide = M(:,1) <  datenum(2019,7,1); % start of July
>> out = M(idb&ide,:)
out =
   737579.72727        7.00000       99.00000       22.00000       56.00000       18.00000       95.00000       67.00000       79.00000
   737582.59091       70.00000       68.00000       33.00000       73.00000       93.00000       87.00000       99.00000       44.00000
   737585.45455       78.00000       57.00000       99.00000       21.00000       42.00000       27.00000       13.00000       22.00000
   737588.31818       19.00000       95.00000       79.00000        4.00000       43.00000       67.00000       15.00000       10.00000
   737591.18182       70.00000       33.00000       75.00000       67.00000       97.00000       46.00000       64.00000       89.00000
   737594.04545       42.00000       48.00000       14.00000       18.00000       61.00000       65.00000       26.00000       48.00000
   737596.90909       62.00000        7.00000       68.00000       18.00000        6.00000       76.00000       72.00000       90.00000
   737599.77273       39.00000       77.00000       59.00000       79.00000       83.00000       77.00000       98.00000       48.00000
   737602.63636       90.00000       93.00000       84.00000        4.00000       84.00000       66.00000       42.00000       21.00000
   737605.50000       74.00000       13.00000       66.00000       12.00000        1.00000       74.00000       64.00000       48.00000

And checking the first column to see that it only contains dates from June:

>> datestr(out(:,1))
ans =
03-Jun-2019 17:27:16
06-Jun-2019 14:10:54
09-Jun-2019 10:54:32
12-Jun-2019 07:38:10
15-Jun-2019 04:21:49
18-Jun-2019 01:05:27
20-Jun-2019 21:49:05
23-Jun-2019 18:32:43
26-Jun-2019 15:16:21
29-Jun-2019 12:00:00