Task: Create a csv file with all the secondary sleep windows

childmindresearch / app-actigraphy

A webapp for annotating actigraphy data

GNU Lesser General Public License v3.0

0 stars 0 forks source link

Task: Create a csv file with all the secondary sleep windows #70

Closed nathaliaesper closed 8 months ago

nathaliaesper commented 8 months ago

Description

Since the app supports multiple sleep windows per day, it would be good to have a csv file with all the secondary sleep windows. For example, if we flag a day with two sleep windows, the first one should be added to the sleeplog_X file, and the second one should be added to a new file called multiple_sleep_X, where X is the subject number. The content of this file should be: the first column is the date, the second column is the sleep onset time, and the third column is the sleep offset time (wake up).

Date; sleep_onset1; wakeup_1; sleep_onset2; wakeup_2
21-02-2024; 14:34; 15:50; 18:00; 18:30

Tasks

[ ] Get the input from all the sleep windows (except for the first one) and add them to the CSV file.

Freeform Notes

No response

ReinderVosDeWael commented 8 months ago

What should be filled in for the days that do not have the maximum number of sleep windows?

ReinderVosDeWael commented 8 months ago

@nathaliaesper This format can be quite ambiguous; for example:

If the only sleep time is after midnight, and we anchor it to the prior day akin to how the UI works, then the time will show up with the wrong date.
However we place the date, anything that crosses midnight will have the wrong date for one of the times.
I'll take the liberty to add the timezone ;).

In any case, I'll need a default value; should it conform to the GGIR 'fake' output for days that have no timestamps (currently set to Jan-1-1970 00:00:00+00:00, but let me know if you want something else there), should be it be empty/None, the old 3AM default?

Also any reason why you wouldnt want this in a long format like so? It'd reduce some of the ambiguity and ascertain that you don't have to deal with wrangling the missing values in a wide csv.

onset, wakeup
2024-02-21 14:52:16+05:00, 2024-02-21 18:52:16+05:00
2024-02-21 19:52:16+05:00, 2024-02-21 21:52:16+05:00

nathaliaesper commented 8 months ago

Hi @ReinderVosDeWael

Re timezone: sure, I forgot to mention it. The suggested file format (see below) works well for us.

onset, wakeup 2024-02-21 14:52:16+05:00, 2024-02-21 18:52:16+05:00 2024-02-21 19:52:16+05:00, 2024-02-21 21:52:16+05:00

I don't understand the reason for having a default value, but if that is the case, it can be the current set to Jan-1-1970 00:00:00+00:00.

ReinderVosDeWael commented 8 months ago

A defualt value is no longer needed with a long format; this is only necessary in the wide one where some days may have more sleep windows than others.