Extract a portion of character

Hi... I've got a set of characters that I need to extract, convert to numbers, and perform a task on to find the differences between them. Here is an example of data:

'12/04/2022 15:22:52.211.674'

'12/04/2022 15:22:52.211.752'

'12/04/2022 15:22:52.211.830'

'12/04/2022 15:22:52.211.908'

I just need the last 6 numbers, such as below:

211.674

211.752

211.830

211.908

Next, I need to find the difference between the numbers, for example:

211.752 - 211.674

211.830 - 211.752

211.908 - 211.830

and store it in the new column.

Thank you:)

Answer 1

[Walter Roberson]

It would be possible to strip those values down to the milliseconds and nanoseconds as a simple text operation. However I did not do that here, because it is not robust: you need to assume that you might cross a boundary at some point. The code here uses the full precision of the inputs -- in theory it can even handle historic leap seconds.

format long g
D = {
    '12/04/2022 15:22:52.211.674'
    '12/04/2022 15:22:52.211.752'
    '12/04/2022 15:22:52.211.830'
    '12/04/2022 15:22:52.211.908'
};
D2 = regexprep(D, '\.(\d\d\d)$', '$1');  %fraction of seconds must be adjacent
D3 = datetime(D2, 'InputFormat', "MM/dd/yyyy HH:mm:ss.SSSSSS", 'Format',"MM/dd/yyyy HH:mm:ss.SSSSSS" )

D3 = 4×1 datetime array
   12/04/2022 15:22:52.211674
   12/04/2022 15:22:52.211752
   12/04/2022 15:22:52.211830
   12/04/2022 15:22:52.211908

[~, ~, S] = hms(D3)    %if you really want to see the seconds

S = 4×1

52.211674 52.211752 52.21183 52.211908

%but do not look at just the seconds for the calculation, as you might have
%crossed a second boundary
nanodiff = seconds(diff(D3)) * 1e6   %nanoseconds

nanodiff = 3×1

78.000000000003 77.9999999999745 78.0000000000314