These are chat archives for nightscout/intend-to-bolus

8th
Oct 2017
dmdfreak
@dmdfreak
Oct 08 2017 01:14

No smb_enact needed. Temp refreshed: monitor/temp_basal.json: {"duration":30,"rate":1.05,"temp":"absolute"}
No bolus needed (yet). Settings less than 15 minutes old. Edison on battery: 58%. Pumphistory-24 < 30m old
Settings refreshed. RefreshValueError: need more than 0 values to unpack
ed pumphistory and meal.json
IOB: -0.026
Completed supermicrobolus pump-loop at Sat Oct 7 20:06:49 CDT 2017:

Starting supermicrobolus pump-loop at Sat Oct 7 20:07:02 CDT 2017 with 15 second wait_for_silence:
Waiting up to 4 minutes for new BG: ............glucose.json newer than pump_loop_completed
Radio ok. Listening: .No pump comms detected from other rigs
Preflight OK. Profile less than 60m old. Temp duration >20m. Ref
SMB pump-loop failed. Unsuccessful supermicrobolus pump-loop at Sat Oct 7 20:09:52 CDT 2017

Throwing some errors.
Scott Leibrand
@scottleibrand
Oct 08 2017 01:19
are you always getting the same error in the same place? If so, you'll need to drill down on that command
Brian Rabinovitz
@bdr1177
Oct 08 2017 01:21
@scottleibrand Unfortunately, my daughter is away camping right now so I can't dig into the logs.
Scott Leibrand
@scottleibrand
Oct 08 2017 01:23
you might want to set up papertrail for next time
Brian Rabinovitz
@bdr1177
Oct 08 2017 01:24
We're also all of the sudden today getting a bunch of bytearray errors. The rig has needed restarting a few time but has recovered. Just don't know what is causing these errors all of the sudden.
Oh I have papertrail. I didn't realize the log you are referring to is accessible from papertrail.
Scott Leibrand
@scottleibrand
Oct 08 2017 01:53
probably depends on how you have papertrail configured, but if not you should add it
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:03
Question: If BG data is old, say 372, will OpenAps still use this to calculate and deliver a temp? Seems like it did and sent BG into the ground.
image.png
Two 0.6U temps when in target; there's no reason for that.
need to understand what's happening here.
Scott Leibrand
@scottleibrand
Oct 08 2017 02:08
you'll have to look at the OpenAPS pill or the pump-loop.log to see its reasoning, but the 0.6U temps were probably neutral temps. they certainly didn't "send BG into the ground" there, as they were only about 10% above your normal basal. something else was apparently causing BG to drift downward, and the question is whether OpenAPS could've done anything to prevent it, and if so, why didn't it
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:08
Oct 07 18:24:28 DiabeticinJapan pump-loop.log: enact/enacted.json: {"insulinReq":0,"bg":74,"reservoir":null,"temp":"absolute","snoozeBG":82,"rate":0.6,"predBGs":{"IOB":[74,75,75,76,76,76,77,77,77,78,78,78,78,78,79,79,79,79,79,80,80,80,80,80,80,80,80,80,81]},"minPredBG":999,"IOB":-0.05,"reason":"COB: 0, Dev: 4, BGI: 0, ISF: 70, Target: 90, minPredBG 79, IOBpredBG 81; 82-79 in range: no temp required; setting current basal of 0.6 as temp. ","COB":0,"eventualBG":82,"recieved":true,"duration":30,"tick":"+1","timestamp":"2017-10-08T09:53:16.641833","deliverAt":"2017-10-08T00:53:11.176Z"}
Scott Leibrand
@scottleibrand
Oct 08 2017 02:08
for one thing, it looks like you weren't looping very reliably
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:09
What's this log mean then, the bit about "setting current basal of 0.6 as temp."
Scott Leibrand
@scottleibrand
Oct 08 2017 02:10
what's your BG target range?
I'm guessing it's 80-100 or something like that?
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:10
60-190 at the moment
Scott Leibrand
@scottleibrand
Oct 08 2017 02:10
huh what?
your bolus wizard targets?
60-190 would average to 125, and it says your target is 90, so that can't be right
since BG was predicted to get back up to 81 or 82 based on your negative IOB, and that was within the target range, it reported 82-79 in range: no temp required and set a neutral temp.
if you want OpenAPS to intervene earlier, I would recommend raising your bolus wizard min_bg (the low end of your target range) to 100
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:14
Why's there a second temp though?
Scott Leibrand
@scottleibrand
Oct 08 2017 02:15
what do you mean by "a second temp"?
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:16
In the image, there's two 0.6U basals.
Scott Leibrand
@scottleibrand
Oct 08 2017 02:17
probably for similar reasons, but you'd need to look at the reason field to see
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:17
reason field?
Scott Leibrand
@scottleibrand
Oct 08 2017 02:17
in the openaps pill or the pump-loop.log
like you pasted above
from 18:55 or so
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:20
first mention of any temp starts at 17:47: 90, minPredBG 77, IOBpredBG 77; Eventual BG 78 < 80, setting 30m zero temp.
But then this
Target: 90, minPredBG 79, IOBpredBG 81; 82-79 in range: no temp required; setting current basal of 0.6 as temp.
that's what I don't understand
shouldn't that also be a zero
Scott Leibrand
@scottleibrand
Oct 08 2017 02:21
that's because your pump bolus wizard target ranges are too wide, like I said earlier.
82 is in range, so it doesn't want to zero temp
if you don't think 82 is in range, you need to raise the low end of the range to not be at 80
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:22
I'm completely lost.
sorry, really tired :)
Scott Leibrand
@scottleibrand
Oct 08 2017 02:22
go into your pump bolus wizard settings menu and see what it says
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:23
K, will have to wait until friend wakes up :)
Scott Leibrand
@scottleibrand
Oct 08 2017 02:23
or we can check it in the settings files on the pump
cat ~/myopenaps/settings/bg_targets.json
you'll probably see "high": 100, and "low": 80,
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:25
yes
exactly that
"user_preferred_units": "mg/dL", "targets": [ { "i": 0, "high": 100, "start": "00:00:00", "low": 80, "offset": 0, "x": 0 } ], "first": 1
Scott Leibrand
@scottleibrand
Oct 08 2017 02:26
ok, there's the problem. you need to change the pump settings to set the target range to be 100-100 instead of 80-100
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:26
Okay, lemme see if I get this correctly...
Scott Leibrand
@scottleibrand
Oct 08 2017 02:26
and then it will low temp whenever eventualBG is below 100, unless BG is rising faster than expected
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:27
OpenAps said insuling was needed because at the time, the low target was 80, but BG was 89
So openaps said gives insulin to get to 80
yes?
Scott Leibrand
@scottleibrand
Oct 08 2017 02:27
no
it wasn't looking at the current BG
it was looking at the eventualBG, which is where the BG is predicted to end up after IOB takes effect
that was 82, because BG was drifting downward
but since 82 is > 80, it said "everything's fine" and allowed your normal basal of 0.6U/hr to continue running
it wasn't saying "you need extra insulin", just "we have no reason to shut it off"
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:29
so if target was 100, openaps would have said wait, we're gonna hit 82, stop doing the temp.
yes?
Scott Leibrand
@scottleibrand
Oct 08 2017 02:34
correct
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:34
oivay
Ok, will relay.
If we set 120 as high and 100 as low, OpenAPS would wait until 120 before taking action yes?
Scott Leibrand
@scottleibrand
Oct 08 2017 02:42
with 120 high and 100 low, it would low-temp any time eventualBG is below 100, but would not high temp unless eventualBG is above 120. in between it would set neutral temps.
we generally find that setting high and low to the same value works better, as it can keep BGs in a slightly tighter range. but you can start with a 20 point spread and narrow it down over time if you prefer
Deweyoxberg
@Deweyoxberg
Oct 08 2017 02:46
ok; pump settings were updated but the cat command still shows a low target for 80
last sync was 1 minute ago
Deweyoxberg
@Deweyoxberg
Oct 08 2017 03:19
ah, fixed
just slow :D
dmdfreak
@dmdfreak
Oct 08 2017 03:39
Running oref1, have everything in preferences.json marked true and rebooted. A BG of 220 and going up and only 1.79 IOB. I feel like it should have been hitting me with SMBs and nothing is happening.
Scott Leibrand
@scottleibrand
Oct 08 2017 03:39
what does your pump-loop.log say?
dmdfreak
@dmdfreak
Oct 08 2017 03:41
one sec. I just installed papertrail. let me read it
Oct 07 22:40:22 timsedison pump-loop.log: COB: 5 remainingCItotal/csf: 4.3 remainingCarbs: 5
Oct 07 22:40:22 timsedison pump-loop.log: UAM Impact: -4.1 mg/dL per 5m; UAM Duration: 0 hours
Oct 07 22:40:22 timsedison pump-loop.log: minPredBG: 86 minIOBPredBG: 84 minCOBPredBG: 99 minUAMPredBG: 84 avgPredBG: 86 COB: 5 carbs: 80
Oct 07 22:40:22 timsedison pump-loop.log: BG projected to remain above 80 for 240 minutes
Oct 07 22:40:22 timsedison pump-loop.log: bgUndershoot: -61 zeroTempDuration: 240 zeroTempEffect: 265 carbsReq: -63
Oct 07 22:40:23 timsedison pump-loop.log: Checking deliverAt: 2017-10-08T03:40:22.852Z is within 1m of current time: Sat Oct 7 22:40:23 CDT 2017
Oct 07 22:40:23 timsedison pump-loop.log: and that smb-suggested.json is less than 1m old
Oct 07 22:40:23 timsedison pump-loop.log: enact/smb-suggested.json: {"insulinReq":0,"bg":208,"reservoir":"57.0","temp":"absolute","snoozeBG":141,"predBGs":{"COB":[208,196,185,175,166,157,149,141,135,128,123,118,114,111,108,106,104,102,102,101,101,100,100,100,99,99,99,99,99,99,99,99,99,100,100,101,101,102,102,103,104,104,105,106,106,107,107,108],"aCOB":[208,205,200,194,188,180,172,164,157,151,145,140,136,132,129,126,124,122,121,120,119,118,117,116,115,114,113,113,112,111,110,110,109,109,108,108,108,108,107,107,107,107,107,108],"UAM":[208,196,185,175,165,156,148,141,134,127,122,117,112,109,105,103,100,98,97,96,95,94,94,93,92,91,90,89,88,88,87,86,86,85,85,85,84],"IOB":[208,196,185,175,165,156,148,141,134,127,122,117,112,109,105,103,100,98,97,96,95,94,94,93,92,91,90,89,88,88,87,86,86,85,85,85,84]},"IOB":1.604,"reason":"COB: 5, Dev: 25, BGI: -7.92, ISF: 57, Target: 80, minPredBG 86, minGuardBG 84, IOBpredBG 84, COBpredBG 108, UAMpredBG 84; IOB 1.6 > max_iob 1, temp 0.95 ~ req 0.95U/hr. ","COB":5,"eventualBG":141,"tick":-12,"deliverAt":"2017-10-08T03:40:22.852Z"}
Oct 07 22:40:25 timsedison pump-loop.log: No smb_enact needed. Temp refreshed: monitor/temp_basal.json: {"duration":21,"rate":0.95,"temp":"absolute"}
Oct 07 22:40:40 timsedison pump-loop.log: No bolus needed (yet). Settings less than 15 minutes old. Edison battery at 66% is charged (>= 98%) or likely charging (60-70%). pumphistory-24h refreshed
Oct 07 22:40:41 timsedison pump-loop.log: Completed supermicrobolus pump-loop at Sat Oct 7 22:40:41 CDT 2017:
Oct 07 22:41:02 timsedison pump-loop.log: Starting supermicrobolus pump-loop at Sat Oct 7 22:41:02 CDT 2017 with 2 second wait_for_silence:
dmdfreak
@dmdfreak
Oct 08 2017 03:49
Reading the numbers, it says it should come down to in range. which is set to 85-90 in the pump.
Dana Lewis
@danamlewis
Oct 08 2017 03:50
Side note - You have a maxiob of 1, is that intentional?
Also yes, it's predicting you to come down with your current level of iob.
dmdfreak
@dmdfreak
Oct 08 2017 03:52
It was set to 0 before but I thought I read that it should be set to the same as your general hourly basel rate
Should I put it back to 0?
Dana Lewis
@danamlewis
Oct 08 2017 03:55
If it's set to 0, OpenAPS will only high temp to get net iob back to 0, then no more. With a maxiob of 1 (this is total units; not a rate of per hour), it will only ever give you an extra unit of insulin. I would re-read the docs, there's a helpful chart in there now too to illustrate when various settings may limit (or not) the dosing:
You should put it at what you feel comfortable with and what you want it to do
dmdfreak
@dmdfreak
Oct 08 2017 03:57
ok. Then I originally had it set to zero before and I was comfortable with it. I just couldn't remembered. I'll go read that now.
Dana Lewis
@danamlewis
Oct 08 2017 03:59
Just know at max iob 0, it's essentially only going to prevent lows, and slight help with minor high temps to recover. That's essentially predictive low glucose suspend only, a la a 640. Having that number > 0 turns it so it can help with both highs and lows. A la 670. You should set it however you want; but most people start at 0 for only a few days until they understand the system, and then adjust from there.
dmdfreak
@dmdfreak
Oct 08 2017 04:04
I think I left it at zero before I updated and never changed it because it was giving me SMBs all the time. I put it back to zero and I'm going to go read that doc.
Dana Lewis
@danamlewis
Oct 08 2017 04:09
+1 for reading the docs. You may also want to consider turning SMB's off if you are unhappy with them and would prefer temps only.
dmdfreak
@dmdfreak
Oct 08 2017 04:14
I was liking the SMBs. I would prefer to leave them on. My problem was that I wasn't seeing any this time around. Even now I'm watching the logs and it still isnt giving me any SMBs with the logs predicting UAM 168 and IOB 193
Dana Lewis
@danamlewis
Oct 08 2017 04:18
It's not going to SMB or high temp you with a maxiob at 0
dmdfreak
@dmdfreak
Oct 08 2017 04:23
OH ok I just got it..... Geez.... I'm sorry I had it correctly set before I updated but I forgot to check my settings before updating and lost them. And I have to relearn what I'm looking at everytime because I wait to long and I'm old LOL
Immediately started working again as soon as I set it to 4.
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:04

image.png
@scottleibrand Hi Scott! Thanks for helping @YYCMichael earlier about my OpenAPS giving me more insulin. I see you got confused because I'm using a non-standard OpenAPS system. Unfortunately CGM that is supported by OpenAPS is not available in Japan, so I have to work with Libre FCM (Flash Glucose Monitoring) instead of CGM. That's why I'm not looping "reliably".

My question here would be why OpenAPS decided to give me insulin when my BG was only 92.

Dana Lewis
@danamlewis
Oct 08 2017 05:05
Same responses as before - need to see your logs! It's not just about bg currently, but your level of insulin activity and where your BG is predicted to be, relative to your target.
Deweyoxberg
@Deweyoxberg
Oct 08 2017 05:06
@danamlewis : any log in specific? I can get the details we need to troubleshoot.
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:07
any log in specific?
I assume you mean the log for when OpenAPS triggered the treatment?
Deweyoxberg
@Deweyoxberg
Oct 08 2017 05:08
Oct 07 18:12:12 DiabeticinJapan pump-loop.log: enact/enacted.json: {"insulinReq":0,"bg":74,"reservoir":null,"temp":"absolute","snoozeBG":82,"rate":0.6,"predBGs":{"IOB":[74,75,75,76,76,76,77,77,77,78,78,78,78,78,79,79,79,79,79,80,80,80,80,80,80,80,80,80,81]},"minPredBG":999,"IOB":-0.05,"reason":"COB: 0, Dev: 4, BGI: 0, ISF: 70, Target: 90, minPredBG 79, IOBpredBG 81; 82-79 in range: no temp required; setting current basal of 0.6 as temp. ","COB":0,"eventualBG":82,"recieved":true,"duration":30,"tick":"+1","timestamp":"2017-10-08T09:53:16.641833","deliverAt":"2017-10-08T00:53:11.176Z"}
Dana Lewis
@danamlewis
Oct 08 2017 05:08
Pump-loop log. Which you can get by typing l when you're logged into the rig, or by looking at the OpenAPS pill in NS, or via papertrail as you did before
Deweyoxberg
@Deweyoxberg
Oct 08 2017 05:09
That's when the basal started
Dana Lewis
@danamlewis
Oct 08 2017 05:09
Keep in mind a "neutral temp" means it's setting a temp basal rate at what your underlying basal is. So that's not giving you "extra" insulin, because you're predicted to be in range.
Setting neutral temps is by design, to make it easier to see that you are looping successfully.
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:10

Ahhh, okay so this means even though it looks like a temp basal- it's actually just setting what it believed my basal would be by default as a temp basal?

Keep in mind a "neutral temp" means it's setting a temp basal rate at what your underlying basal is. So that's not giving you "extra" insulin, because you're predicted to be in range.

Dana Lewis
@danamlewis
Oct 08 2017 05:11
Yes, exactly.
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:11
that makes sense. I understand!
Dana Lewis
@danamlewis
Oct 08 2017 05:12
If you're running Autotune and it's using a slightly different underlying basal and you have not updated NS, it might be visualized as more or less insulin . So, if that trips your brain up, might be worth updating your NS profile to match Autotune.
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:12
I'd love to- where can I find what Autotune is using?
Dana Lewis
@danamlewis
Oct 08 2017 05:12
But you'll probably be ok now that you know what "neutral" temps are doing by seeing that in the logs.
Pop into your Autotune folder in the rig and look at the Autotune-recommendations
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:13
Okay, thanks! We'll have a look!
Dana Lewis
@danamlewis
Oct 08 2017 05:14
:+1:
Deweyoxberg
@Deweyoxberg
Oct 08 2017 05:14
That's in the myopenaps directory yes?
Think we found it: 07:30 | |
08:00 | | 0.595
08:30 | |
09:00 | | 0.590
09:30 | |
10:00 | | 0.522
.590 :D
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:18
That answers the question. :)
Dana Lewis
@danamlewis
Oct 08 2017 05:18
:+1:
(And then basals get set at however the pump can round to)
Deweyoxberg
@Deweyoxberg
Oct 08 2017 05:19
@danamlewis : for the NS settings, if the recommendation is a blank row, what do I enter for that time?
Dana Lewis
@danamlewis
Oct 08 2017 05:21
That's half hour increments - don't bother with those
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:29
Hey @danamlewis - Is it possible to tell Autotune about the limitations of my pump? For example- My pump can only do 0.05 increments, so I can't do 0.36, only 0.35
Dana Lewis
@danamlewis
Oct 08 2017 05:30
Nope. Autotune will tune as needed, and then the pump will round accordingly based on its limitations
#DiabeticinJapan
@LuminaryXion_twitter
Oct 08 2017 05:31
Darn, was hoping to make it a bit easier for me to update with ^_^;
Thanks!
LilDucky
@LilDucky
Oct 08 2017 05:36
With the insulin absorption curves in 0.6.0, are they active without oref1 featured?
Dana Lewis
@danamlewis
Oct 08 2017 05:40
They are independent of any other feature if you enable them
LilDucky
@LilDucky
Oct 08 2017 05:57
Ah, cool. Thanks @danamlewis. it is tempting to go to 0.6.0 just for that - we are still too new for SMB.
Scott Leibrand
@scottleibrand
Oct 08 2017 06:05
if you don't enable any advanced features, the 0.6.0-dev changes are fairly minor: main differences are around predicted carb absorption for AMA, and better autotune (which will likely be a patch release here soon as well)
LilDucky
@LilDucky
Oct 08 2017 06:08
@scott, thanks. We are having some issues which I think are related to the "tail" of the bilinear calcs. May look at AMA sooner rather than later too.
Scott Leibrand
@scottleibrand
Oct 08 2017 06:16
AMA is on by default: if you enter carbs, it will help out as needed with temps
that's way more powerful with SMBs, of course, but still helpful without
LilDucky
@LilDucky
Oct 08 2017 06:22
👍
Tore Bjørndalen
@BjoTore_twitter
Oct 08 2017 06:46
Last nights autotune. I'm obviously getting much more resistent, but recommendation is restricting the change. What should I do?
Oct 08 00:15:10 ToresEdison autotune.log: oldCR: 4.896 fullNewCR: 3.178 newCR: 4.552
Oct 08 00:15:10 ToresEdison autotune.log: Limiting adjusted ISF of 7.63 to 20.57 (which is pump ISF of 28.8 / 1.4 )
Oct 08 00:15:10 ToresEdison autotune.log: p50deviation: 2.73 p50BGI -3.6 p50ratios: 0.18 Old ISF: 42.373 fullNewISF: 7.627 adjustedISF: 20.571 newISF: 38.013
Oct 08 00:15:10 ToresEdison autotune.log: Autotune pump profile recommendations:
Oct 08 00:15:10 ToresEdison autotune.log: ---------------------------------------------------------
Oct 08 00:15:10 ToresEdison autotune.log: Recommendations Log File: /root/myopenaps/autotune/autotune_recommendations.log
Oct 08 00:15:10 ToresEdison autotune.log: /usr/local/bin/oref0-autotune-recommends-report: line 22: warning: setlocale: LC_NUMERIC: cannot change locale (en_US.UTF-8): No such file or directory
Oct 08 00:15:14 ToresEdison autotune.log: Parameter | Pump | Autotune
Oct 08 00:15:14 ToresEdison autotune.log: -------------------------------------
Oct 08 00:15:14 ToresEdison autotune.log: ISF [mg/dL/U] | 28.800 | 38.013
Oct 08 00:15:14 ToresEdison autotune.log: Carb Ratio[g/U]| 4.000 | 4.552
Oct 08 00:15:14 ToresEdison autotune.log: Basals [U/hr] | - |
Oct 08 00:15:14 ToresEdison autotune.log: 00:00 | 0.500 | 0.608
Oct 08 00:15:14 ToresEdison autotune.log: 01:00 | | 0.690
Oct 08 00:15:14 ToresEdison autotune.log: 02:00 | | 0.693
Oct 08 00:15:14 ToresEdison autotune.log: 03:00 | | 0.700
Oct 08 00:15:14 ToresEdison autotune.log: 04:00 | 0.700 | 0.847
Oct 08 00:15:14 ToresEdison autotune.log: 05:00 | | 0.930
Oct 08 00:15:14 ToresEdison autotune.log: 06:00 | | 0.902
Oct 08 00:15:14 ToresEdison autotune.log: 07:00 | 1.000 | 1.042
Oct 08 00:15:14 ToresEdison autotune.log: 08:00 | | 0.922
Oct 08 00:15:14 ToresEdison autotune.log: 09:00 | | 0.897
Oct 08 00:15:14 ToresEdison autotune.log: 10:00 | | 0.904
Oct 08 00:15:14 ToresEdison autotune.log: 11:00 | 0.800 | 0.739
Oct 08 00:15:14 ToresEdison autotune.log: 12:00 | | 0.722
Oct 08 00:15:14 ToresEdison autotune.log: 13:00 | | 0.722
Oct 08 00:15:14 ToresEdison autotune.log: 14:00 | | 0.664
Oct 08 00:15:14 ToresEdison autotune.log: 15:00 | 0.900 | 0.853
Oct 08 00:15:14 ToresEdison autotune.log: 16:00 | | 0.930
Oct 08 00:15:14 ToresEdison autotune.log: 17:00 | | 0.963
Oct 08 00:15:14 ToresEdison autotune.log: 18:00 | | 0.955
Oct 08 00:15:14 ToresEdison autotune.log: 19:00 | | 0.935
Oct 08 00:15:14 ToresEdison autotune.log: 20:00 | | 0.945
Oct 08 00:15:14 ToresEdison autotune.log: 21:00 | | 0.938
Oct 08 00:15:14 ToresEdison autotune.log: 22:00 | 1.100 | 1.066
Oct 08 00:15:14 ToresEdison autotune.log: 23:00 | | 1.096
What's the fullNewCR and fullNewISF numbers?
"Limiting adjusted ISF of 7.63 to 20.57 (which is pump ISF of 28.8 / 1.4 )" seems extreme,
Tore Bjørndalen
@BjoTore_twitter
Oct 08 2017 06:55
image.png
I guess that due to my high ISF, SMB's are being limited by MinGuardBG all the time? I went skyhigh.
This was yesterday during lunch.
Tore Bjørndalen
@BjoTore_twitter
Oct 08 2017 07:03
I've been on FIASP since sep 20th and I understand that this may be variations that can happen.
Scott Leibrand
@scottleibrand
Oct 08 2017 07:47
when did autotune raise your ISF? looks like it's trying to lower it, but it'll take a few days starting from that high...
if you want it to adjust more quickly, you can change the autosens min and/or max and re-run autotune to get it to follow those, and then widen them again to autotune from there...
Tore Bjørndalen
@BjoTore_twitter
Oct 08 2017 08:59
since the 3rd it has been rising and then suddenly turned. I'll look into the quicker version. Thanks.
What's the fullNewCR and fullNewISF numbers?
Scott Leibrand
@scottleibrand
Oct 08 2017 09:40
the raw calculated CR or ISF based on just that day's data, before any adjustments for safety caps etc.
Tore Bjørndalen
@BjoTore_twitter
Oct 08 2017 11:04
I see, again thanks, your support is amazing!
fidoman
@efidoman
Oct 08 2017 11:38
Screenshot from 2017-10-08 07-37-59.png
The most recent week 10/1-10/8 I've been using eSMB. The thing that stands out as different is the 0% low which is the biggest plus for me. I need to avoid lows more than anything else. However, eSMB keeps everything else in line with my needs/expectations too. Awesome!
@scottleibrand - I will try entering carbs at the time I start eating dinner as well and see how it goes.
fidoman
@efidoman
Oct 08 2017 11:46
Below is last night where I had 60g for dinner but only announced 15g, then I announced the other 45g later on and a couple of bedtime snacks a bit afterwards. eSMB did great. Maybe would have been even better if I announced the full 60g at dinner time.
Screenshot from 2017-10-08 07-45-07.png
Steve Bernacki
@sbernacki
Oct 08 2017 12:04
@scottleibrand thanks for the clarification g4-local v g5. I guess I meant g5 receiver. When I plug it into the rig, I do see the kernel picking the receiver as a USB device. Are there any specific log entries that help indicate it's pulling from the receiver? I'll plug it all in and try again :)
JamOrHam
@jamorham
Oct 08 2017 12:30
Is anyone who is looping interested in responding to a public radio journalist (Dr. Cheryl K. Olson) looking for people in the NYC or SF area to talk about artificial pancreas and its implications etc? If so PM me and I can send over the contact email address.
cameronrenwick
@cameronrenwick
Oct 08 2017 12:46
hey need some quick help. I have a guy who's offered to sell me a dexcom G4 receiver, model mt-20649-2 for $250. This looks all on the up and up but I just want to make sure. Any thoughts on this or what I should be checking? I've got my G4 running and collecting data on my xdrip+ so I'll be able to check the dexcom receiver when I meet up with the guy... just don't wanna drop $250 and regret
Diabetic_Dude
@diabetic_dude_twitter
Oct 08 2017 12:54
I can across this article this morning and I plan on signing up learn programming pay what you want specifically I’m interested in their section on JavaScript
Diabetic_Dude
@diabetic_dude_twitter
Oct 08 2017 13:04
@cameronrenwick I suppose it all depends on your need. That version is at least two versions back. Behind the G4+share and G5.
Steve Bernacki
@sbernacki
Oct 08 2017 13:19
So we just set up some IFTTT buttons for eating/activity/cancel temp basal. Running the eating / activity, we can see the bar show up on the Nightscout portal, but when we run the cancel temp basal the bar for the existing temp basal doesn't disappear. However, looking at the pump-loop logs, I see that her target is back to normal. Maybe I just need to upgrade our Nightscout instance? Running 0.9.2 currently.
Steve Bernacki
@sbernacki
Oct 08 2017 13:32
Update.. if we refresh the browser, the bar shortens to indicate the duration it was on for. So looks like just a visual glitch. Curious how Matt managed to make the treatment go completely away when he selected cancel treatment.. actually deleting the record versus just shortening it, perhaps? https://www.youtube.com/watch?v=0ck23JTa2Wk&feature=youtu.be
Marcus Lindfors
@Mackwe
Oct 08 2017 14:16
Strange thing that has started recently: Using OpenAPS on Edison and for some reason, the OpenAPS pill could say "Temp basal started, 0.35 for 30 m" followed by "enacted". The NS basal pill still says T: 0U. The basal line says 0u. Refreshing the NS page doesn't change any of it. What can I do to prevent this?
jaylagorio
@jaylagorio
Oct 08 2017 15:13
@sbernacki I offer you no solutions, just confirmation - I've never had a temp target bar disappear when doing a cancel temp unless I refresh the browser for at least the last year or so.
I can't explain what's going on in the video
cameronrenwick
@cameronrenwick
Oct 08 2017 15:25
@diabetic_dude_twitter thanks for that info. I have G4 and right now am sticking with it... whether I move to a G5 I don't know. I don't use share aspects at all (don't know if that's one of the features that work in Canada?). I really just want to be able to do offline looping between my rig and pump
Tim Street
@tim2000s
Oct 08 2017 16:26
@Mackwe Have you updated to the latest NS version?
live4sw
@live4sw
Oct 08 2017 16:26
When you are having a low, how do you input the rescue carbs? I am generally just inputting 15g on my pump with bolus wizard and bolusing 0.1U so that it registers on NS, but the problem is that when it is fast-acting carbs, they aren't decaying quickly enough and so SMB boluses on the way up, causing me to yo-yo back down.
My basals and ratio are based on autotune and I'm pretty sure they're ok otherwise.
Tim Street
@tim2000s
Oct 08 2017 16:28
@live4sw Set a high temp target at the same time. That way you stop SMB.
I believe @danamlewis has an IFTTT button that enters carbs + high temp target in one hit.
live4sw
@live4sw
Oct 08 2017 16:30
smart, I am using IFTTT for exercise mode but not using the button for low treatment
sounds like i should switch to inputting the low treatment with a customized IFTTT button like the "Low Treatment with a 60 min high target to help recovery" in the docs
I've yo-yo'ed on two low treatments two days in a row (usually only happens on weekends since i'm walking around so much), so this sounds like a better tactic.
Marcus Lindfors
@Mackwe
Oct 08 2017 16:33
@tim2000s Running 0.10.0-release-20170922.. Is that too old?
Tim Street
@tim2000s
Oct 08 2017 16:34
That should be fine.
Dana Lewis
@danamlewis
Oct 08 2017 16:51
@live4sw yep. Doesn't have to be a full 60 minutes, I think my low treatment is 4 carbs plus a 110 target for 20 minutes and that works well for me. (I also have 120 and 140 targets on my watch, so I can set those later if I want )
live4sw
@live4sw
Oct 08 2017 17:08
wow, YDMV is certainly true! i usually need at least 15g to treat a low, and low-temping often takes an hour or more to meaningfully arrest a downward trend
i'm just an average sized guy with maybe a slightly more active than average lifestyle
Scott Leibrand
@scottleibrand
Oct 08 2017 17:09
@sbernacki there is /var/log/openaps/cgm-loop.log
Dana Lewis
@danamlewis
Oct 08 2017 17:10
@live4sw Well, put in whatever carb amount you want for your most common low treatment. My point is for the high temp to disable smb's, it doesn't have to be 60 minutes. And that's the point of the carbs plus target combo button
Steve Bernacki
@sbernacki
Oct 08 2017 17:11
@scottleibrand I had found that, but mine only contains Sun Oct 8 13:08:02 EDT 2017 AttributeError: 'NoneType' object has no attribute 'iter_records' 60 2017-10-08T12:58:56-04:00 Sun Oct 8 13:09:02 EDT 2017 AttributeError: 'NoneType' object has no attribute 'iter_records' 60 2017-10-08T12:58:56-04:00 Sun Oct 8 13:10:02 EDT 2017 over and over
live4sw
@live4sw
Oct 08 2017 17:11
Ah, gotcha, so a high target will disable SMB for the length of time of the high target?
Dana Lewis
@danamlewis
Oct 08 2017 17:11
@efidoman thx for continuing to share feedback
@live4sw yes. But high as in one point above 100. So a 110 does just fine for disabling SMB's . If you eat carbs and set a 140+ you're likely going to yoyo yourself
You may want that if you're going to be super active. But if goal is just not to SMB on the recovery rise, a lower high target works better from my experience
live4sw
@live4sw
Oct 08 2017 17:13
Got it. Is that true even if "enableSMB_with_bolus" is what I have enabled rather than enableSMB_with_temptarget?
Scott Leibrand
@scottleibrand
Oct 08 2017 17:13
@sbernacki not working then. You can also do openaps use cgm iter_glucose 1 to check it manually. You said lsusb shows the receiver correctly?
What does openaps use cgm config show?
Steve Bernacki
@sbernacki
Oct 08 2017 17:14
Oh this doesn't seem good:
root@edison1:~# lsusb
unable to initialize libusb: -99
Scott Leibrand
@scottleibrand
Oct 08 2017 17:14
@live4sw yes, a >100 temp target disabled SMBs for any reason.
@sbernacki nope. Check your cable etc. try reversing it
Steve Bernacki
@sbernacki
Oct 08 2017 17:15
root@edison1:~/myopenaps# openaps use cgm config
{
  "model": "G5"
}
Dana Lewis
@danamlewis
Oct 08 2017 17:15
@live4sw yes. Disables all SMB for whatever reason temporarily during the time frame
Scott Leibrand
@scottleibrand
Oct 08 2017 17:16
Also check the charging icon on the receiver lights up.
I think all this troubleshooting stuff is in the docs somewhere too.
Steve Bernacki
@sbernacki
Oct 08 2017 17:24
OK this is looking more promising. I actually just unplugged everything and plugged it in but now I'm seeing:
root@edison1:~/myopenaps# openaps monitor-cgm
cgm://JSON/oref0_glucose/raw-cgm/raw-entries.json
reporting raw-cgm/raw-entries.json
tz://JSON/rezone/cgm/cgm-glucose.json
reporting cgm/cgm-glucose.json
I'll experiment with switching out some cables, the OTG cable we have is from an old G4->NS uploader phone setup so it's seen some action.
Nalla306
@Nalla306
Oct 08 2017 18:11
Not sure if this is the right place to ask. I have an issue with night scout displaying information from openaps in the trace. I see the entries appear in the admin panel as future treatments.....but they are time stamped around 6 minutes before the actual time of entry. Deleting them has no effect as they reappear. Time on iPhone,pump and Edison all match. Any suggestions? Thanks
Saskia
@Missdazzle
Oct 08 2017 18:19

Set up a Raspberry Pi from scratch with autosens and API Secret (instead of token) and w/out autotune. Works.
Now I wanted to run autotune manually on the command line. But when I start the command "oref0-autotune...." it says:

ERROR: API_SECRET is not set when calling oref0-autotune.sh

I tried filling up the autotune start command with something like --api_secret=<myAPISecret as it is or hashed> but that hasn´t been the trick :)... What do I have to do instead?

Diadon81
@Diadon81
Oct 08 2017 18:45
Hi, today I start to getting the following error: Old settings refresh at Module.load (module.js:356:32)
ed. pump_loop_completed less than 10m ago. Refresh at Module.load (module.js:356:32)
ed pumphistoryCould not parse input data: [SyntaxError: Unexpected end of input]
and meal.json
I have tried to re-run oref0-runagain, last time I deleted myopenaps directory, but it doesn’t help, any ideas?
Dana Lewis
@danamlewis
Oct 08 2017 18:54
Check for typo in your preferences.json
And if not, start drilling down and running things manually
Diadon81
@Diadon81
Oct 08 2017 18:55
Ok, I will try
Diadon81
@Diadon81
Oct 08 2017 19:26
My fault, it was a mistype in preferences.json
Dana Lewis
@danamlewis
Oct 08 2017 19:58
that's the main issue people are seeing with update, so I added that to the troubleshooting notes of the 0.6.0-dev PR: https://github.com/openaps/oref0/pull/694#issue-263319617
Screen Shot 2017-10-08 at 12.58.03 PM.png
Nalla306
@Nalla306
Oct 08 2017 20:21
Would setting my pump time 6 minutes ahead fix my issue? tried every google fix I could find but all my entries from openaps into night scout from my pump go into the future entries ....but return after removing and aren't actually in the future but around 6 minutes behind the actual time of entry.....driving me insane
Dana Lewis
@danamlewis
Oct 08 2017 20:24
@Nalla306 you could try that - will likely break looping, but it should try to re-set your pump clock for you to the correct time
while you're trying things, make sure your time zone on the rig is correct?
FYI all - just published the autotune changes we've been testing in dev for a while to master so everyone can take advantage of them. oref0 0.5.5 is now master: https://github.com/openaps/oref0/releases
these are pretty significant fixes to autotune, so if you are running master, I would recommend (when you have spare time and brain space) updating to this version of master if you are an autotune user
(if you're on dev, you already have these autotune fixes)
tynbendad
@tynbendad
Oct 08 2017 20:32
ah, i was just about to ask what 0.5.5 was ;)
you guys move faster than i can keep up (which is probably a good thing cuz i'm slow :)
Scott Leibrand
@scottleibrand
Oct 08 2017 20:36
@Missdazzle that error has been downgraded to a warning in 0.5.5. The API_SECRET is only needed if you've locked down your NS to prevent anonymous read access.
If you do want to set it, it expects it to be an environment variable, not a command-line argument.
Steve Bernacki
@sbernacki
Oct 08 2017 20:44

So I wanted to share this comment from my wife, whom I've built our rig for. Words can't express my gratitude to you all @danamlewis , @scottleibrand and everyone who continues to develop this amazing technology. You are truly changing lives. Thank you :)

Words can't describe how "free" I've felt these last few days. For the last 33 years, I've worried about if I've taken enough or too much insulin for what I've ate. This is why, for the most part, I have chosen to eat very similarly every day of my life. I've also worried about going too low at night and not waking up especially when exercising or drinking the day/evening prior. All in all, it's a shit load of "worry" that only people with insulin dependent diabetes can begin to relate to. For those who know me, you know I don't talk about this kind of stuff. In fact, the only time you'll hear me complaining is when my diabetes gets in the way of me living my life! A HUGE WEIGHT has been lifted from my brain (if that makes sense?). I feel so light right now. :) Thank you Nightscout and #OpenAps

live4sw
@live4sw
Oct 08 2017 20:49
Random question - my current daytime target range is 110-110, but this was the last summary in NS. What does "Target: 80" mean?
Screenshot 2017-10-08 at 4.48.59 PM.png
Dana Lewis
@danamlewis
Oct 08 2017 20:52
@sbernacki <3 <3 <3 <3 thank you for sharing with us <3 <3 <3 <3
Nalla306
@Nalla306
Oct 08 2017 20:54
change time notes appear on night scout trace after setting pump 6 minutes ahead.....cob appear in openaps pill but no bolus or carbs on trace....they all appear in future entries in admin tools. checked timezone and its right....entries in future entries are still 6 minutes behind.
I see this error in paper trail at the same timestamps repeated, not sure if releveant
Scott Leibrand
@scottleibrand
Oct 08 2017 20:55
@live4sw you probably have advanced target adjustments (formerly "adjust target when high") enabled in preferences
with that enabled, when your BG is high it will lower the target from your usual setpoint to as low as 80. once your eventualBG is below 80, it will revert back to normal target (and usually zero temp for a bit). this allows it to more aggressively correct highs.
Dana Lewis
@danamlewis
Oct 08 2017 20:57
@live4sw I think we've discussed this before, too. I would re-visit your preferences.json and make sure you understand what all you might have enabled.
Nalla306
@Nalla306
Oct 08 2017 20:57
08 21:50:22 myloop libmraa: edison: Arduino board not detected, assuming Intel Edison Miniboard
Oct 08 21:50:22 myloop libmraa: libmraa initialised for platform 'Intel Edison' of type 2
Oct 08 21:50:26 myloop libmraa: libmraa version v1.7.0 initialised by user 'root' with EUID 0
Oct 08 21:50:26 myloop libmraa: gpio214: init: Failed to write to 'export': No such device
Oct 08 21:50:26 myloop libmraa: edison: Arduino board not detected, assuming Intel Edison Miniboard
Oct 08 21:50:26 myloop libmraa: libmraa initialised for platform 'Intel Edison' of type 2
Oct 08 21:50:30 myloop libmraa: libmraa version v1.7.0 initialised by user 'root' with EUID 0
Scott Leibrand
@scottleibrand
Oct 08 2017 20:58
@Nalla306 that's normal: Intel's Edison software doesn't know about the Explorer Board
live4sw
@live4sw
Oct 08 2017 20:58
Thanks @danamlewis and @scottleibrand. Will go back and reread that part of the docs and recheck my preferences.json
Nalla306
@Nalla306
Oct 08 2017 21:03
@scottleibrand cheers...that was only odd looking thing in paper trail at the same time stamp. Bolus wizard entries make their way to night scout but all end up in future entries but at a timestamp 6 minutes earlier than the actual entry. Times on iPhone, rig and pump all match, along with timezone on rig. I must have been trying for about 5 hours now with every fix I could find but no solution. Infuriating!
Dana Lewis
@danamlewis
Oct 08 2017 21:04
also, I found a bug in github this week. fun! Tl;dr - the 0.6.0-dev PR keeps moving, it's now here: openaps/oref0#707
(if you go to an old link and it looks closed, head to https://github.com/openaps/oref0/pulls and look for the open 0.6.0-dev PR). I keep updating the content in the latest open one
Saskia
@Missdazzle
Oct 08 2017 21:15
@scottleibrand Okay, right, env shows no API Secret. And how do I get it there?
Scott Leibrand
@scottleibrand
Oct 08 2017 21:18
export API_SECRET=yourlongplaintextsecret is one way. totally not necessary though if you upgrade to 0.5.5: then you can just ignore the warning
Saskia
@Missdazzle
Oct 08 2017 21:21
Well that´s the version I´m already on.
Dana Lewis
@danamlewis
Oct 08 2017 21:21
did you just update?
to what I just released 30 min ago?
you're probably on 0.5.4, what we had before that if you were seeing that, because we hadn't released 0.5.5 from your first post
Saskia
@Missdazzle
Oct 08 2017 21:23
No, I didn´t update in the last few minutes. But npm view oref0 version shows me a 0.5.5 ...... - strange
Saskia
@Missdazzle
Oct 08 2017 21:37
Thanks @danamlewis and @scottleibrand . I need to do some more reading about autotune. I didn´t get yet where to look for the results. :)
spaldings
@spaldings
Oct 08 2017 21:38

Now getting Nightspot Boot Error

Unable to setup authorization
{"name":"MongoError","message":"not master and slaveOk=false","stack":"MongoError: not master and slaveOk=false
at Function.MongoError.create (/app/node_modules/mongodb/node_modules/mongodb-core/lib/error.js:31:11)
at queryCallback (/app/node_modules/mongodb/node_modules/mongodb-core/lib/cursor.js:200:36)
at Callbacks.emit (/app/node_modules/mongodb/node_modules/mongodb-core/lib/topologies/server.js:119:3)
at null.messageHandler (/app/node_modules/mongodb/node_modules/mongodb-core/lib/topologies/server.js:358:23)
at Socket. (/app/node_modules/mongodb/node_modules/mongodb-core/lib/connection/connection.js:292:22)
at Socket.emit (events.js:95:17)
at Socket. (_streamreadable.js:765:14)
at Socket.emit (events.js:92:17)
at emitReadable
(_stream_readable.js:427:10)
at emitReadable (_stream_readable.js:423:5)","ok":0,"errmsg":"not master and slaveOk=false","code":13435,"codeName":"NotMasterNoSlaveOk"}

Any thoughts?

Brian Rabinovitz
@bdr1177
Oct 08 2017 22:31
Is there still a problem with calculating COB after CGM warmup? This doesn't look quite right to me.
IMG-1661 (1).PNG
IMG-1662.PNG
This is on 0.6.0 dev from yesterday.
Dana Lewis
@danamlewis
Oct 08 2017 22:32
possibly @bdr1177 - we just tested some fixes last night to further help this , if you want to grab cob-interpolationbranch and see if that changes your calculations
openaps/oref0#703
(checkout & pull that branch; bash runagain; update your preferences, etc.)
Brian Rabinovitz
@bdr1177
Oct 08 2017 22:35
@danamlewis Okay, I'll try it.
Brian Rabinovitz
@bdr1177
Oct 08 2017 23:01
@danamlewis Yes, that appeared to work. Carbs have dropped down to what looks like a reasonable number.
Dana Lewis
@danamlewis
Oct 08 2017 23:01
cool. if you'd like to thumbs up that PR, we can get that into dev for further testing
Brian Rabinovitz
@bdr1177
Oct 08 2017 23:03
Will do.
Brian Rabinovitz
@bdr1177
Oct 08 2017 23:10
@danamlewis I gave the pr a thumbs up, but the description of the pr seems like the opposite of the problem I was describing (description says it's to fix carbs decaying to zero too quickly).
Dana Lewis
@danamlewis
Oct 08 2017 23:34
FYI all - just merged exercise mode into dev (0.6.0) for future testing. Make sure you read the notes before you consider enabling this variable:
Screen Shot 2017-10-08 at 4.33.48 PM.png
Scott Leibrand
@scottleibrand
Oct 08 2017 23:35
@bdr1177 depending on the timing, the bug could result in erroneously high COB values too. I'll update the PR to note that as well
updated and merged. thanks for the help testing.