Woot! Built my third rig this morning before work. Took about an hour. Many impressive improvements in the setup script and docs since my last build. I have a deep appreciation and respect for all of you! Many, many thanks!
:+1:
@parenthetic :)
:clap:
@Kdisimone I was getting what looked to trend longer than 16 hours on 4400 mah lipo, i think, on dev branch with: a) underclocking to 500 MHz; b) Pi Zero W LEDs off ...
but it would be good for me to do a few clean passes at reproducing this data
not bad
i’m building it the way anyone following the openaps docs would
@mhaeberli @Kdisimone are you both referering to the same dev branch for the battery test?
0.7.0-dev or dev?
I am - I ran 0.7.0-dev about a week ago
I may have mis-spoken about the loop size; I’ll confirm when I get home tonight
That is, I may have been using the 6600
@alimhassam plain dev branch. I was testing that out since that’s the branch that is likely to be released soon. Just trying to test things out like new pi users would be encountering.
:+1: that's what it sounded like so i wanted to point out that you both seemed to be refering to battery life on a different branch :-) from my understanding it should get better on 0.7.0-dev.
Testing the 0.7.0-dev on Edison/EXB on 4400mAh battery this weekend - will let people know
Autotune is showing a need to almost halve my carb ratio on 0.7.0-dev for me. I've tried lowering it slightly and the recommendation keeps getting lower. Also, if I do go lower (around 16) then I start having to feed potential post-meal lows, so I bump carb ratio back up. Any ideas/suggestions?
Parameter | Pump | Autotune
-------------------------------------
ISF [mg/dL/U] | 52.000 | 43.247
Carb Ratio[g/U]| 20.000 | 10.000
Basals [U/hr] | - |
00:00 | 0.800 | 0.765
01:00 | | 0.739
02:00 | | 0.744
03:00 | 0.900 | 0.774
04:00 | 0.900 | 0.807
05:00 | | 0.811
06:00 | 0.750 | 0.622
07:00 | 0.600 | 0.558
08:00 | | 0.527
09:00 | | 0.551
10:00 | | 0.589
11:00 | 0.750 | 0.574
12:00 | | 0.665
13:00 | | 0.675
14:00 | | 0.679
15:00 | 0.700 | 0.644
16:00 | | 0.638
17:00 | 0.750 | 0.661
18:00 | | 0.715
19:00 | | 0.732
20:00 | | 0.732
21:00 | | 0.735
22:00 | | 0.763
23:00 | | 0.766
The other point is that I moved my basals down to this current level about a week ago they were average 0.90 and the current pump basals are close to those recommendations. Now it is suggesting further lowering of basals. Same lowering for ISF. A week ago it was 60 and recommendation was about 50 (I think), so I lowered it to 52. Now, it is recommending ISF of 43.
@efidoman you’ll need to look at the log files for each day’s run to see why it’s recommending lowering things.
Yesterday's log showed the following:
Limiting CSF to 4.11 (which is 1.5 * pump CSF of 2.736842105263158 )
totalMealCarbs: 0 totalDeviations: 0 oldCSF 5.345 fullNewCSF: 5.345007782966422 newCSF: 4.105
Limiting fullNewCR from 8.854 to 9.50 (which is 0.5 * pump CR of 19 )
Limiting CR to 9.50 (which is 0.5 * pump CR of 19 )
oldCR: 8.994 fullNewCR: 9.5 newCR: 9.5
Limiting adjusted ISF of 8.94 to 34.67 (which is pump ISF of 52 / 1.5 )
p50deviation: 3.27 p50BGI -4.33 p50ratios: 0.186 Old ISF: 48.073 fullNewISF: 8.942 adjustedISF: 34.667 newISF: 45.392
Looks like I'm getting a lot of negative numbers for csf and BGI but I'm not sure how to interpret it
1 mealCOB: 50.4 mealCarbs: 135 basalBGI: 2.7 BGI: -19.9 IOB: 7.6 at 15:22:13 dev: 11.93 avgDelta: -8.00 csf
1 mealCOB: 48.1 mealCarbs: 135 basalBGI: 2.7 BGI: -19.5 IOB: 7.1 at 15:27:12 dev: 12.27 avgDelta: -7.25 csf
1 mealCOB: 46.0 mealCarbs: 135 basalBGI: 2.7 BGI: -19.1 IOB: 6.6 at 15:32:13 dev: 11.31 avgDelta: -7.75 csf
1 mealCOB: 44.2 mealCarbs: 135 basalBGI: 2.7 BGI: -18.6 IOB: 6.2 at 15:37:13 dev: 9.81 avgDelta: -8.75 csf
1 mealCOB: 42.3 mealCarbs: 135 basalBGI: 2.7 BGI: -18.0 IOB: 5.8 at 15:42:13 dev: 9.75 avgDelta: -8.25 csf
1 mealCOB: 40.1 mealCarbs: 135 basalBGI: 2.7 BGI: -17.4 IOB: 5.4 at 15:47:13 dev: 11.93 avgDelta: -5.50 csf
1 mealCOB: 37.3 mealCarbs: 135 basalBGI: 2.7 BGI: -16.8 IOB: 5.0 at 15:52:13 dev: 14.80 avgDelta: -2.00 csf
1 mealCOB: 63.9 mealCarbs: 165 basalBGI: 2.7 BGI: -16.1 IOB: 4.7 at 15:57:13 dev: 18.15 avgDelta: 2.00 csf
1 mealCOB: 60.0 mealCarbs: 165 basalBGI: 2.7 BGI: -15.7 IOB: 6.6 at 16:02:13 dev: 20.99 avgDelta: 5.25 csf
1 mealCOB: 55.8 mealCarbs: 165 basalBGI: 2.7 BGI: -15.7 IOB: 6.6 at 16:07:13 dev: 22.42 avgDelta: 6.75 csf
1 mealCOB: 51.6 mealCarbs: 165 basalBGI: 2.7 BGI: -15.5 IOB: 6.2 at 16:12:13 dev: 22.28 avgDelta: 6.75 csf
1 mealCOB: 47.9 mealCarbs: 165 basalBGI: 2.7 BGI: -15.4 IOB: 6.1 at 16:17:13 dev: 20.11 avgDelta: 4.75 csf
1 mealCOB: 44.6 mealCarbs: 165 basalBGI: 2.7 BGI: -15.1 IOB: 5.7 at 16:22:12 dev: 17.62 avgDelta: 2.50 csf
1 mealCOB: 41.9 mealCarbs: 165 basalBGI: 2.7 BGI: -14.8 IOB: 5.3 at 16:27:12 dev: 14.56 avgDelta: -0.25 csf
1 mealCOB: 39.7 mealCarbs: 165 basalBGI: 2.7 BGI: -14.4 IOB: 5.0 at 16:32:12 dev: 11.70 avgDelta: -2.75 csf
1 mealCOB: 37.7 mealCarbs: 165 basalBGI: 2.7 BGI: -14.0 IOB: 4.6 at 16:37:13 dev: 10.54 avgDelta: -3.50 csf
1 mealCOB: 36.1 mealCarbs: 165 basalBGI: 2.7 BGI: -13.6 IOB: 4.3 at 16:42:12 dev: 8.58 avgDelta: -5.00 csf
during that timeframe, you had enough IOB that, without the carb impact, BG would have been falling very rapidly, but as the deviations increased to be even higher than the BGI, the BG stopped dropping and started rising from 15:57 - 16:22 before dropping again
That makes sense and reflects what happened. I'll have to set aside some weekend time to dig in further. Thanks.
you might want to do some manual calculations of total insulin delivered for a few meals (probably from NS reservoir values) vs. carbs given for that meal, and total basal insulin that would've been delivered over that timeframe, to see what CR "really" should be.
I have Oref1 installed on my main rig with autotune running manually. It is issuing SMBs. However, on my secondary rig, which has the same settings, although I have run autotune manually, SMBs are still disabled - my log shows the following message: Warning: Autotune has not been run. All microboluses will be disabled until you manually run autotune or add it to run nightly in your loop. Is there another step that I need to take after running a manual autotune to enable SMB?
it is looking for a ~/myopenaps/autotune/ directory
if you're running autotune somewhere else, you can just
mkdir that directory to make it SMB
Thanks @scottleibrand . I ran the mkdir -p ~/newdirectory/settings && cp ~/myopenaps/settings/profile.json ~/newdirectory/settings/autotune.json && cp ~/myopenaps/settings/pumpprofile.json ~/newdirectory/settings/pumpprofile.json command to set up the manual autotune. Do I need to run that again or does the snippet you suggested do something different?
this is different: just
mkdir -p ~/myopenaps/autotune/
(the steps you did was to run it in a separate directory, so OpenAPS won't use those settings, also in case you didn't realize that)
Thanks @scottleibrand . I have run the command now. @danamlewis yeah - it wasn't my first choice: I started with autotune running automatically every night but after some weeks of trying, I have been unable to stay in range in the afternoon with a single I:C ratio and single ISF. On a typical workday, my I:C and ISF requirements are out by at least a factor of two compared to the rest of a 24hr cycle. (I:C of 1:5 at lunch, ISF of 1.4mmol through the afternoon vs I:C of 1:10, or 1:11 and ISF of 3.5). Even with these more aggressive settings through lunch/afternoon and SMBs enabled, I tend to be in the 160-200 range.
fair - just wanted to make sure you understood why you have to create the directory separately from running autotune since you did it that way
thanks @danamlewis understanding is important - and being able to give the system the commands so it is set up correctly is definitely a learning curve for me. I appreciate the support!
:+1:
@scottleibrand - thanks for the suggestion of manual CR calculations. Will do.