Sampling architecture #81
Conversation
| ## TODO: For future | ||
| #data_sampling: | ||
| ## Possible methods: time, count, fraction | ||
| ## starttime and endtime format is dd-mm-yyyy hh:mm:ss in UTC timezone |
There was a problem hiding this comment.
Would it be useful to be able to specify an array of ranges? This way if you wanted a value single range you could just specify that and if you wanted a sequence of ranges this could also be provided.
There was a problem hiding this comment.
Good point, will add that.
| # count: 2 | ||
| #method: userid | ||
| #config: | ||
| # userids: |
There was a problem hiding this comment.
Do we want to keep fixed on subjectID or userID? For historical reasons, SubjectID is the name we chose for the main ID we use on the platform (with potentially UserID being introduced later when we have the self-enrollment portal). It may not result in too much confusion, as this is more on the analysis side, but I'd point this out perhaps it is sensible to be consistent.
There was a problem hiding this comment.
could change it to subjectId. I used user-id because the ID column in the output files is key.userID
|
@afolarin added the code for multiple time ranges. Please let me know if that looks good to you. |
Added user and data sampling mechanisms. User sampling mechanism contains option to choose users by fraction, count and IDs Data sampling mechanisms include choosing data between time ranges, by count and by fraction.