Option(s) to Enable Raw Data #386
Conversation
|
Looks good Jason... great for mobile because it renders faster than IOB-COB. |
|
With the last commits the ENABLE option is requred |
|
+1. Also tested with ENABLE array of 'careportal rawbg' |
|
@ELUTE / @YYGIRL, basic raw data in /pebble now looks like: I'm using the same field names we use elsewhere, might not be what you had when using the uploader from @hackingtype1's fork {
"status": [{"now":1422602605616}],
"bgs":[{"sgv":"164","bgdelta":-9,"trend":5,"direction":"FortyFiveDown","datetime":1422602506000,"battery":"87","filtered":200960,"unfiltered":189472,"rssi":185}],
"cals":[{"_id":"54caf5305b6d4ba395ba0b3a","device":"dexcom","date":1422587207000,"dateString":"Thu Jan 29 19:06:47 PST 2015","slope":1006.8263237320385,"intercept":30000,"scale":1,"type":"cal"}]
} |
|
TY Jason |
…an be used to get more data
… options correctly
|
ready for final testing and review |
|
A few notes on possible UI/UX for raw data |
|
Is it possible to display noise? I think that would be helpful as well. |
|
@ELUTE maybe as a tooltip?, but the raw dots are so small might be hard to get the hover events on them. |
|
At the very least, there should be no "raw" dot drawn if the noise level is clean. |
|
The minor difference that we see while the noise is clean is only because we're using a simplified formula? |
|
Yes, omitting white dots when Dex is using raw to calculate EGV makes sense - limits confusion. Also makes a clear distinction to viewer that the noise level is or is not clean, and whether the displayed value (receiver) is now based on the filtered value. |
|
I always want to see the white dot, even if the noise level is clean. Also note that we have people using ped and 505 receivers, which is going Plus what are you going to show when clean and in a special value? You have I want more data, not less. On Sun, Feb 1, 2015 at 11:46 AM, John Costik [email protected]
Join the Jack Attack! |
|
Right now while correcting a mild low noise is clean, dex shows 67, but raw is 76. I can't think of a reason I don't want to see that 76. |
|
There have been too many times where the raw helped me feel more confident On Sun, Feb 1, 2015 at 11:54 AM, Jason Calabrese [email protected]
Join the Jack Attack! |
|
If the noise level is "clean" then the white dot is merely our generalized calculation based on what's we know of Dexcom's. The value Dexcom is showing IS based on the raw value. The only time the noise is clean w/ a special value is when the special value is 5 (start up cal period). It will be very risky to show folks data that maps to EGV during this period, particularly with new sensors. Restarted sensors are a bit easier, as we can use the last know slope & intercept to calculate the EGV. If we want to show people the raw values during this period w/ new sensors, I would remove any and all numbers from the graph. |
Option(s) to Enable Raw Data
|
I understand. But the raw value shows a disturbing trend much sooner than On Sun, Feb 1, 2015 at 1:15 PM, John Costik [email protected]
Join the Jack Attack! |
|
If the Dexcom is reporting clean noise, then your calculations based on raw values should have a consistent offset from any displayed "clean" data. There should be no need to show a secondary calculation based on the same raw value, certainly not one that is not fully tested and verified. -John
|
|
This pull request was merged to master a little early, we still need to come to some consensus on the display of raw data when the noise level is clean. My proposal is to do that in a new pull request before the |
|
My bad - I have no idea why the codehub app triggered the merge. -John
|
Goal is a have a lab that will allow more advanced users to enable raw data in the mainline version
New
ENABLEOption:ENABLE="rawbg"orENABLE="rawbg-on"(to default to on)Tasks:
wip/iob/cobbranchENABLEenv varcalrecords and filtered, unfiltered, and rssi rawbg fields in/pebbleendpoint