search

ddediu / AdhereR

Computation of adherence to medications from Electronic Healthcare Data in R
25 stars 6 forks source link

Show carryover in sliding windows and per episode plots ? #172

Closed ddediu closed 2 years ago

ddediu commented 2 years ago

Investigate if it is desirable to show the carry over boxes also in the per episode and sliding window plots (as they are in the simple CMA plots). Possibly provide a parameter to explicitly force them to be shown (by default FALSE to keep the current behavior of not showing them)?

ddediu commented 2 years ago

Seems to require a pretty massive restructuring of the way sliding windows and per episodes are computed, as the necessary info is simply not stored right now, so the potting function does not have the info...

ddediu commented 2 years ago

... on a second thought, I think there might be a way of recovering that info using some data.table tricks :) let me see...

dffyfe commented 2 years ago

I can send tomorrow, if you make process.patient return a list then do data[, process.patient(.SD)$listitem1, by=Id.colname] then run it for the CMA and event.info. The event info will be there for each window I.D but using I. In data.table to get max gives you the original

7 Nov 2021 22:13:02 Dan Dediu @.***>:

... on a second thought, I /think/ there might be a way of recovering that info using some data.table tricks :) let me see...

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub[https://github.com/ddediu/AdhereR/issues/172#issuecomment-962689797], or unsubscribe[https://github.com/notifications/unsubscribe-auth/ARVBNWWMCGBSIILRB372MDDUK32W5ANCNFSM5HOAVHGQ]. Triage notifications on the go with GitHub Mobile for iOS[https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675] or Android[https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub]. [data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAD8AAAA/CAYAAABXXxDfAAAAAXNSR0IArs4c6QAAAARzQklUCAgICHwIZIgAAAAmSURBVGiB7cEBDQAAAMKg909tDwcUAAAAAAAAAAAAAAAAAAAAJwY+QwABivJx1AAAAABJRU5ErkJggg==###24x24:true###][Tracking image][https://github.com/notifications/beacon/ARVBNWVGXVZKMUZC6DD6RBTUK32W5A5CNFSM5HOAVHG2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOHFQXWBI.gif]

ddediu commented 2 years ago

ok. I was thinking about making it return all the info and then subset it to extract the event.info columns -- I dont know which is faster, might need some benchmarking

dffyfe commented 2 years ago

I was getting a weird data.table error saying I could only return a list or atomic vector using by= when I tried that, even though I was literally trying to return a list.

7 Nov 2021 22:23:04 Dan Dediu @.***>:

ok. I was thinking about making it return all the info and then subset it to extract the event.info columns -- I dont know which is faster, might need some benchmarking

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub[https://github.com/ddediu/AdhereR/issues/172#issuecomment-962691427], or unsubscribe[https://github.com/notifications/unsubscribe-auth/ARVBNWSL7UK36PXV2OYZ57LUK334PANCNFSM5HOAVHGQ]. Triage notifications on the go with GitHub Mobile for iOS[https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675] or Android[https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub]. [data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAD8AAAA/CAYAAABXXxDfAAAAAXNSR0IArs4c6QAAAARzQklUCAgICHwIZIgAAAAmSURBVGiB7cEBDQAAAMKg909tDwcUAAAAAAAAAAAAAAAAAAAAJwY+QwABivJx1AAAAABJRU5ErkJggg==###24x24:true###][Tracking image][https://github.com/notifications/beacon/ARVBNWRAHIRGNAF5SE7Y3YLUK334PA5CNFSM5HOAVHG2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOHFQYCYY.gif]

ddediu commented 2 years ago

I did it in a different way: I am merging the two in a single data.table, return it, and then subset them again outside -- seems to work pretty well :)

ddediu commented 2 years ago

ok, works and it adds a bit of overhead but it's acceptable (by default it is disabled so it has to be requested explicitly for plotting). It seems I might have a working solution, ow it's just to find time to implement it :)

dffyfe commented 2 years ago

Hi Dan,

I think I've got committing to GitHub slightly figured out (and got around working on a networked computer). Definitely easier than using 7zip every time!

I've put it here for now as I'm still getting my head round it and don't want to mess anything up.

dffyfe/AdhereRFork: Fork of AdhereR to be used in STU/MAVIS project (github.com) https://github.com/dffyfe/AdhereRFork

Best,

David

On Sun, Nov 7, 2021 at 10:49 PM Dan Dediu @.***> wrote:

ok, works and it adds a bit of overhead but it's acceptable (by default it is disabled so it has to be requested explicitly for plotting). It seems I might have a working solution, ow it's just to find time to implement it :)

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/ddediu/AdhereR/issues/172#issuecomment-962695335, or unsubscribe https://github.com/notifications/unsubscribe-auth/ARVBNWXDJCKDG4IKNW6RGSTUK367TANCNFSM5HOAVHGQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

ddediu commented 2 years ago

done for sliding windows; to implement for per episodes and in the Shiny interface...

ddediu commented 2 years ago

Should be fixed in devel