You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
That is, it should be possible to take two event frames, extract from them a set of compatible fields (compatible in the sense of having the same types and names – possibly after renaming) and then combine these into a single event frame which contains all the rows from both.
This is the equivalent of SQL's UNION operation and similar to Pandas concat().
I don't think this would be an enormous piece of work to implement, but it wouldn't be trivial and no doubt there will be annoying edge cases which Hypothesis will find for us. Possibly the trickiest bit will be designing a sensible API.
An example of the sort of thing which could only reasonably be done with this feature would be counting the number of episodes for each patient where the events that form part of an episode could be recorded in either of two tables.
That is, it should be possible to take two event frames, extract from them a set of compatible fields (compatible in the sense of having the same types and names – possibly after renaming) and then combine these into a single event frame which contains all the rows from both.
This is the equivalent of SQL's
UNION
operation and similar to Pandasconcat()
.I don't think this would be an enormous piece of work to implement, but it wouldn't be trivial and no doubt there will be annoying edge cases which Hypothesis will find for us. Possibly the trickiest bit will be designing a sensible API.
An example of the sort of thing which could only reasonably be done with this feature would be counting the number of episodes for each patient where the events that form part of an episode could be recorded in either of two tables.
Slack thread:
https://bennettoxford.slack.com/archives/C04DVD1UQC9/p1711373571001209
The text was updated successfully, but these errors were encountered: