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

23rd
Jun 2016
Garrett Webb
@garetis
Jun 23 2016 02:13
When we run cron, I assume we lead it to our loop alias, but are those located in a specific location for the file reference?
Bill Stackpole
@billstackpole
Jun 23 2016 02:20
@garetis - you will need your path and shell specified in crontab
Plus the directive to run every "n", a cd to your openAPS directory
And a call to your OpenAPS loop alias
Does that help?
Garrett Webb
@garetis
Jun 23 2016 02:23
@billstackpole Not at the moment. I'm reading this (https://help.ubuntu.com/community/CronHowto) and trying to reconcile that with what you are saying
Garrett Webb
@garetis
Jun 23 2016 02:29
@billstackpole I'm not sure I know what you mean with the shell reference. If I had to guess this is how I'd set it up right now in my crontab: */10 * * * * cd /home/pi/openaps2 && oref0-reset-git && openaps loop, and I'd access my user crontab by crontab -e
jordan-berger
@jordan-berger
Jun 23 2016 02:33
I am getting the following error when trying to add a dexcom G5: dexcom_reader.constants.CrcError: Could not parse G5EGVRecord
Bill Stackpole
@billstackpole
Jun 23 2016 02:47
@garetis - if you "echo $PATH" at the command line your full path will be returned. At the top of your crontab file (that you edit with the crontab -e command and list with crontab -l) you will need two lines in order for the crontab file to be able to execute your loop.
One line will be "PATH=<your full path from the output of echo $PATH>"
Second will be SHELL=<absolute path to your shell>. (Your shell likely /bin/bash)
Reason is that cron does not have access to either of those so you need to specify it.
Have to run but if u PM me I'll post top of my crontab when I get a min. Should be some discussion in gitter to this end as well if you are willing to search.
Garrett Webb
@garetis
Jun 23 2016 02:51
Ok, Thanks Bill
fallon225
@fallon225
Jun 23 2016 03:20
Is the insulin left in pump alarm working for anyone? W ran out of insulin today, I didn't get any pushover notifications, but I do have them in azure.
Paul Martin
@Spazholio
Jun 23 2016 03:22
@bewest @scottleibrand I’m poking around in MongoDB a bit. In my enacted.json I might see something like this: "reason": "COB: 0, Dev: 2, BGI: -2.08, ISF: 14.6, Target: 89; Eventual BG 87<89, Snooze BG 87, 27m@-1.370 = -0.617 < req -0.23-0.2U”. But I’m not seeing that same amount of info in Mongo. Does it get stripped/parsed out by NS when it gets uploaded? I wanted to try to pull some ISF values out and I’m having no luck so far...
japrobo
@japrobo
Jun 23 2016 03:23
Hey everyone, trying to invoke enact/enacted.json report for the first time and struggling. Should this work? report was created like this: openaps report add enact/enacted.json text oref0 shell determine-basal
ryanchen26
@ryanchen26
Jun 23 2016 03:24
Real quick: lithium AAA in the pump, reading 1.37v, change it or leave it? This is day 9. We'll probably change it...just wondering what others are doing.
Paul Martin
@Spazholio
Jun 23 2016 03:27
@ryanchen26 Personally, that’s when I’d change.
ryanchen26
@ryanchen26
Jun 23 2016 03:28
I shall reassign it to a TV remote. Thanks.
Scott Leibrand
@scottleibrand
Jun 23 2016 03:28
@ryanchen26 I think our threshold is 1.36 or 1.35, and that gives us 8-12h.
Dana Lewis
@danamlewis
Jun 23 2016 03:28
@ryanchen26 there's a note in the docs somewhere with alarm suggestion levels for lithium
Scott Leibrand
@scottleibrand
Jun 23 2016 03:28
@japrobo that looks more like how enact/suggested.json should be configure.
*configured
@Spazholio which collection are you in?
japrobo
@japrobo
Jun 23 2016 03:29
we got enact/suggested.json working well, can invoke it and get good reports
Dana Lewis
@danamlewis
Jun 23 2016 03:29
@japrobo for troubleshooting, consider pasting your output/error message, too
Paul Martin
@Spazholio
Jun 23 2016 03:29
@scottleibrand devicestatus.
Scott Leibrand
@scottleibrand
Jun 23 2016 03:30
@Spazholio and what does the reason field look like in mongo?
Paul Martin
@Spazholio
Jun 23 2016 03:30
Eventual BG 94>90 but Delta -1 < Exp. Delta -0.4; no temp to cancel
Screen Shot 2016-06-22 at 11.32.56 PM.png
Here’s what the whole thing looks like, if that helps.
Marianne Smith, PMP
@mariannesmith77_twitter
Jun 23 2016 03:35

sharing code for @japrobo question. This is what it looks like. I'm sure there is something simple but we can't figure it out.

pi@raspberrypi:~/myopenaps $ openaps report invoke enact/suggested.json
determine-basal://text/shell/enact/suggested.json
{"delta":-4,"glucose":186,"avgdelta":-1.667}
{"duration":0,"rate":0.925,"temp":"absolute"}
{"iob":0,"activity":0,"bolussnooze":0,"basaliob":0,"netbasalinsulin":0,"hightempinsulin":0}
{"max_iob":2,"type":"current","dia":3,"skip_neutral_temps":false,"current_basal":0.95,"max_daily_basal":1.1,"max_basal":2,"min_bg":110,"max_bg":190,"sens":36}
Setting neutral temp basal of 0.95U/hr
reporting enact/suggested.json
pi@raspberrypi:~/myopenaps $ openaps report add enact/enacted.json text oref0 shell determine-basal
added oref0://text/shell/enact/enacted.json
pi@raspberrypi:~/myopenaps $ openaps report invoke enact/enacted.json
oref0://text/shell/enact/enacted.json
/usr/local/bin/oref0-determine-basal iob.json currenttemp.json glucose.json profile.json [[--auto-sens] autosens.json] [meal.json]

Options:
--auto-sens, -a Auto-sensitivity configuration [default: true]
--help Show help [boolean]

reporting enact/enacted.json

Scott Leibrand
@scottleibrand
Jun 23 2016 03:36
@Spazholio that is from May 12. Look at more recent ones since you upgraded?
@mariannesmith77_twitter @japrobo the input to enacted.json report is the suggested.json (the output of determine-basal)
Marianne Smith, PMP
@mariannesmith77_twitter
Jun 23 2016 03:46
We can see that the report is created and then invoked, but cat comes up empty see below
i@raspberrypi:~/myopenaps $ openaps report add enact/enacted.json text oref0 shell monitor/suggested.json
added oref0://text/shell/enact/enacted.json
pi@raspberrypi:~/myopenaps $ openaps report invoke enact/enacted.json
oref0://text/shell/enact/enacted.json
reporting enact/enacted.json
pi@raspberrypi:~/myopenaps $ cat enact/enacted.json
japrobo
@japrobo
Jun 23 2016 03:47
oops, just caught an error. hold that thought, ignore that question for a moment
Scott Leibrand
@scottleibrand
Jun 23 2016 03:48
The point of enacted.json is to actually enact a temp on the pump. How would you go about doing that with openaps use? Once you have it working as a use, you can make a report to do that and capture the output.
japrobo
@japrobo
Jun 23 2016 03:51

ok, we created those reports and got data @scottleibrand

pi@raspberrypi:~/myopenaps $ openaps report invoke enact/set-temp-basal.json
pump://JSON/set_temp_basal/enact/set-temp-basal.json
reporting enact/set-temp-basal.json
pi@raspberrypi:~/myopenaps $ cat set-temp-basal.json
cat: set-temp-basal.json: No such file or directory
pi@raspberrypi:~/myopenaps $ cat enact/set-temp-basal.json
{
"bg": 176,
"temp": "absolute",
"snoozeBG": 156,
"recieved": true,
"rate": 0.925,
"reason": "Eventual BG 156>110 but Avg. Delta -3.33 < Exp. Delta -0.2; setting current basal of 0.95 as temp. Setting neutral temp basal of 0.95U/hr",
"eventualBG": 156,
"timestamp": "2016-06-22T23:49:48.177940",
"duration": 30,
"mealAssist": "Off: Carbs: undefined Boluses: undefined ISF: 36, Target: 150 Deviation: -20 BGI: 0",
"tick": -1

The part we are struggling with is how to input this data into the enact/enacted report?

Scott Leibrand
@scottleibrand
Jun 23 2016 03:53
Start with openaps use -h
Garrett Webb
@garetis
Jun 23 2016 03:54
Hey, does anyone have any input on how to determine my SHELL extension to I can get my crontab operational?
Scott Leibrand
@scottleibrand
Jun 23 2016 03:54
figure out which device you want to use to enact a temp on the pump. That is the main goal: tracking the output is a side effect.
@garetis not sure that question makes sense. Can you expand on what you're trying to do?
Garrett Webb
@garetis
Jun 23 2016 03:55
From Bill's comment earlier...
@garetis - if you "echo $PATH" at the command line your full path will be returned. At the top of your crontab file (that you edit with the crontab -e command and list with crontab -l) you will need two lines in order for the crontab file to be able to execute your loop.
One line will be "PATH=<your full path from the output of echo $PATH>"
Second will be SHELL=<absolute path to your shell>. (Your shell likely /bin/bash)
Reason is that cron does not have access to either of those so you need to specify it.
Scott Leibrand
@scottleibrand
Jun 23 2016 03:55
@japrobo once you figure out the name of the device you want to use to enact the temp, add that with another -h to see all the options for commands that device can run.
I don't think you have to specify your shell.
Garrett Webb
@garetis
Jun 23 2016 03:56
But I do need to specify my path?
Scott Leibrand
@scottleibrand
Jun 23 2016 03:57
Yes. Just type echo $PATH to see your current one.
(Like Bill said) :-)
Garrett Webb
@garetis
Jun 23 2016 03:57
Yeah, I did that and got it already. If I had to guess at this point, this is how I'd set it up right now in my crontab: */10 * * * * cd /home/pi/openaps2 && oref0-reset-git && openaps loop
Scott Leibrand
@scottleibrand
Jun 23 2016 03:58
not sure if you want an && after reset-git. That might work, might not. A ; is safer.
Garrett Webb
@garetis
Jun 23 2016 03:58
do I make /home/pi/openaps2 my full path?
I guess my main question is how do I list the filename/extension so that it points to my loop alias?
Scott Leibrand
@scottleibrand
Jun 23 2016 03:59
You also need to direct the output somewhere, either to logger or directly redirected to a log file, so you can see what cron is doing.
That is what the openaps loop part does.
If you have the PATH set above it so it knows where openaps is.
japrobo
@japrobo
Jun 23 2016 04:00

@scottleibrand, we looked in both of those help files. came up with this use command and got good results, verified on the pump that the command went through

pi@raspberrypi:~/myopenaps $ openaps use pump set_temp_basal enact/suggested.json
{
"bg": 184,
"temp": "absolute",
"snoozeBG": 190,
"recieved": true,
"rate": 2.0,
"reason": "Eventual BG 190>=190, adj. req. rate:3.2 to maxSafeBasal:2.0, temp 0.925<2.0U/hr",
"eventualBG": 190,
"timestamp": "2016-06-22T23:58:57.347028",
"duration": 30,
"mealAssist": "Off: Carbs: undefined Boluses: undefined ISF: 36, Target: 150 Deviation: 6 BGI: 0",
"tick": "+2"

Garrett Webb
@garetis
Jun 23 2016 04:00
Oh, replace the && with ;, I didn't get that earlier
Scott Leibrand
@scottleibrand
Jun 23 2016 04:00
Search Gitter web interface for logger to see how to do the output redirection.
Garrett Webb
@garetis
Jun 23 2016 04:01
No problem. Quick question... what's the gitter web interface? Is that the magnifying glass/search in the top left of this page?
Scott Leibrand
@scottleibrand
Jun 23 2016 04:01
@japrobo perfect. So your report needs to do the same thing: run pump set_temp_basal and store the output to enacted.json.
Web interface means not the app.
The actual Gitter.im website is the only place you can really search. Yes, via that magnifying glass. Or Google.
Garrett Webb
@garetis
Jun 23 2016 04:02
Oh, ok. I didn't know there was an app, I only use this on my laptop. Alright, that's probably enough for me to chew on, thanks.
Scott Leibrand
@scottleibrand
Jun 23 2016 04:03
Google often searches Gitter better than Gitter does.
japrobo
@japrobo
Jun 23 2016 04:04
@scottleibrand, thank you, we got it last night but did not understand how to connect that to enact/enacted.json. we were just calling it something else, made up our own report name and then got all wrapped up trying to attach our new report to enact/enacted, but we already had it figured out and called something else; groan
Scott Leibrand
@scottleibrand
Jun 23 2016 04:05
@garetis are you in ATL still as implied by your GA tech icon? :-)
Garrett Webb
@garetis
Jun 23 2016 04:06
I grew up there, went to undergrad there, but am now in Dallas
Scott Leibrand
@scottleibrand
Jun 23 2016 04:06
Ah
Garrett Webb
@garetis
Jun 23 2016 04:06
Starting MBA at SMU this fall
Garrett Webb
@garetis
Jun 23 2016 04:15
Does this look about right for crontab? */n * * * * cd /home/pi/openaps && oref0-reset-git ; openaps loop > /home/pi/openaps/cron.log 2>&1
Scott Leibrand
@scottleibrand
Jun 23 2016 04:16
n needs to be a number.
Garrett Webb
@garetis
Jun 23 2016 04:16
yeah
and I've set my path
do I need to reboot?
Scott Leibrand
@scottleibrand
Jun 23 2016 04:17
That should work. No reboot, just check that log file after n minutes.
Oh replace > with >>
otherwise it will truncate every time.
Garrett Webb
@garetis
Jun 23 2016 04:17
Noted, thank you. Perfect -- I'm quite excited. Will update tomorrow.
jordan-berger
@jordan-berger
Jun 23 2016 04:27
In the Glucose Data section of the walkthrough,
When i run: openaps report add raw-cgm/glucose-raw.json JSON cgm oref0_glucose --hours 2.0
it returns: added cgm://JSON/oref0_glucose/raw-cgm/glucose-raw.json
When i run: openaps use tz rezone --date dateString --date display_time raw-cgm/glucose-raw.json
It returns:
Traceback (most recent call last):
File "/usr/local/bin/openaps-use", line 4, in <module>
import('pkg_resources').run_script('openaps==0.1.5', 'openaps-use')
File "build/bdist.linux-armv7l/egg/pkgresources/_init.py", line 719, in run_script
File "build/bdist.linux-armv7l/egg/pkgresources/_init.py", line 1504, in run_script
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 63, in <module>
app( )
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/cli/init.py", line 51, in call
self.run(self.args)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/EGG-INFO/scripts/openaps-use", line 57, in run
output = app(args, self)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/init.py", line 92, in call
return self.method.selected(args)(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/init.py", line 31, in call
return self.method(args, app)
File "/usr/local/lib/python2.7/dist-packages/openaps-0.1.5-py2.7.egg/openaps/uses/use.py", line 45, in call
output = self.main(args, app)
File "/usr/local/lib/python2.7/dist-packages/openapscontrib-0.0.15-py2.7.egg/openapscontrib/timezones/_init.py", line 88, in main
inputs = self.get_program(args)
File "/usr/local/lib/python2.7/dist-packages/openapscontrib-0.0.15-py2.7.egg/openapscontrib/timezones/_init.py", line 60, in get_program
program = json.load(argparse.FileType('r')(params.get('input')))
File "/usr/lib/python2.7/argparse.py", line 1142, in call
raise ArgumentTypeError(message % (string, e))
argparse.ArgumentTypeError: can't open 'raw-cgm/glucose-raw.json': [Errno 2] No such file or directory: 'raw-cgm/glucose-raw.json'
Scott Leibrand
@scottleibrand
Jun 23 2016 04:28
See the last line? That is the real error.
You added a report and never invoked it I bet.
jordan-berger
@jordan-berger
Jun 23 2016 04:30
I do not see a command in the instructions to invoke the report.
jordan-berger
@jordan-berger
Jun 23 2016 04:41
Ok, i found the openaps invoke report command.
the "openaps use tz rezone --date dateString --date display_time raw-cgm/glucose-raw.json" runs fine.
the next line in the instructions indicates: "openaps tz rezone --date dateString --date display_time raw-cgm/glucose-raw.json"
but return an error: 'tz' is not an openaps command.
Scott Leibrand
@scottleibrand
Jun 23 2016 04:42
You missed the device add step?
jordan-berger
@jordan-berger
Jun 23 2016 04:43
the openaps device add tz timezones, step?
Section 'device "tz"' already exists
Scott Leibrand
@scottleibrand
Jun 23 2016 04:45
Oh, you just left out use.
Maybe the docs need fixed if they say to do it that way.
jordan-berger
@jordan-berger
Jun 23 2016 04:46
that worked. Not in the docs.
Scott Leibrand
@scottleibrand
Jun 23 2016 04:47
Someone screwed up a merge in that page.
The HEAD and ====== etc shouldn't be there, and there's duplicate content.
Want to clean up and PR a fix?
jordan-berger
@jordan-berger
Jun 23 2016 04:48
I am taking notes and want to redo the instructions for a novice like me.
Not sure what a "PR" is.
Scott Leibrand
@scottleibrand
Jun 23 2016 04:49
Pull request. Did you do the first-pr thing to add your name to the list yet?
Dana Lewis
@danamlewis
Jun 23 2016 04:50
:) pR early and often! There's a my first pr guide in the resources section of docs
Scott Leibrand
@scottleibrand
Jun 23 2016 04:51
@jordan-berger I highly recommend submitting lots of small changes. If you wait until you have everything perfect the docs will already have changed.
One PR per day is probably about right. :-)
jordan-berger
@jordan-berger
Jun 23 2016 04:51
ok, i will read it, and learn how to PR
Dana Lewis
@danamlewis
Jun 23 2016 04:51
:+1:
jordan-berger
@jordan-berger
Jun 23 2016 05:00
Is there a table of reports anywhere in the documentation with a description?
The reports are what is throwing me.
Dana Lewis
@danamlewis
Jun 23 2016 05:04
@jordan-berger no table, but check out the understanding section that was just moved above the walk through sections. @bewest has some good guidance in there on reports etc
Paul Dickens
@thebookins
Jun 23 2016 05:08
Having a problem with status-upload alias as described in the docs (Phase 1). I can run openaps report invoke monitor/upload-status.json followed by ns-upload $NIGHTSCOUT_HOST $NIGHTSCOUT_API_SECRET devicestatus.json monitor/upload-status.json and the data appear in NS no problem. However when I run the alias defined in .ini as status-upload = ! bash -c "openaps report invoke monitor/upload-status.json && ns-upload $NIGHTSCOUT_HOST $NIGHTSCOUT_API_SECRET devicestatus.json monitor/upload-status.json" I get the error "Input file entries.json does not exist"
It looks like it's not picking up the env variables from the alias
anyone seen this before?
Paul Dickens
@thebookins
Jun 23 2016 07:09
Update: looks like I wasn't exporting $NIGHTSCOUT_HOST in ~/.profile after running autoconfigure-device-crud. Will investigate futher. Works now.
Paul Dickens
@thebookins
Jun 23 2016 07:35
Now I've noticed that the dev version of the docs is quite different to the master. I'll go through http://openaps.readthedocs.io/en/dev/docs/walkthrough/phase-1/visualization.html#nightscout-integration carefully later. In the new docs you don't need the env variables at all?
Nathan Morse
@morsen
Jun 23 2016 09:53
@billstackpole Thanks! In the setup you describe, is the pump also the cgm? We’re using Dexcom, and I ended up modifying my glucose report to be like this: “openaps use cgm oref0_glucose —hours 0.5” and use that for input to determine-basal AND ns-upload entries.json.

So, to all, I’m trying to modify my glucose report so that it will fill in network connectivity gaps automatically for upload to nightscout. So, ignoring the timestamp being based on the wrong file currently, I’m trying to get this command to work as an OpenAPS bash-based alias:

SECONDS_SINCE_GLUCOSE=$(expr `date +%s` - $(date -r monitor/glucose.json +%s || expr `date +%s` - 1800)); HOURS_SINCE_GLUCOSE=$(echo 'scale=2; '$SECONDS_SINCE_GLUCOSE' / 3600' | bc); echo $HOURS_SINCE_GLUCOSE && openaps use cgm oref0_glucose --hours $HOURS_SINCE_GLUCOSE > monitor/glucose.json

This command works fine as a standalone input into a shell prompt, but if I include it in openaps.ini like this:

! bash -c "SECONDS_SINCE_GLUCOSE=$(expr `date +%s` - $(date -r monitor/glucose.json +%s || expr `date +%s` - 1800)); HOURS_SINCE_GLUCOSE=$(echo 'scale=2; '$SECONDS_SINCE_GLUCOSE' / 3600' | bc); echo $HOURS_SINCE_GLUCOSE && openaps use cgm oref0_glucose --hours $HOURS_SINCE_GLUCOSE > monitor/glucose.json”

I get this error when running openaps:

  File "/usr/lib/python2.7/ConfigParser.py", line 732, in _interpolate_some
    "'%%' must be followed by '%%' or '(', found: %r" % (rest,))
ConfigParser.InterpolationSyntaxError: '%' must be followed by '%' or '(', found: '%s` - $(date -r monitor/glucose.json +%s || expr `date +%s` - 1800)); HOURS_SINCE_GLUCOSE=$(echo \'scale=2; \'$SECONDS_SINCE_GLUCOSE\' / 3600\' | bc); echo $HOURS_SINCE_GLUCOSE && openaps use cgm oref0_glucose --hours $HOURS_SINCE_GLUCOSE > monitor/glucose.json"'

Can anyone possibly offer some advice? Am I going about this all wrong?

THANKS!!!

Nathan Morse
@morsen
Jun 23 2016 10:04
Ok, it’s totally way too late. Literally all I had to do was quote the percent like it said I should. In my command in openaps.ini, any occurrence of ‘%’ became ‘%%’, and it works. Sorry for the noise, but maybe this info will help someone someday.
Dana Lewis
@danamlewis
Jun 23 2016 11:52
@morsen %? Temp basals need to be in units
Garrett Webb
@garetis
Jun 23 2016 14:29
what I look at my cron log, it only keeps the last 4 loop cycles. Is that typical?
Sarah Hodson Grady
@shgrady
Jun 23 2016 14:31
any suggestions on what percentage your max basal is over your standard basal, mine is only a few 0.1ths over my typical 1.3? Probably a question for my doctor....but how to phrase....
Garrett Webb
@garetis
Jun 23 2016 14:34
I think the docs suggested setting your max basal to 130% of what you currently had set in your pump...
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 14:35
My regular basal is .4. My max is 2. Started with 1 and then increased. It's working well for me.
Garrett Webb
@garetis
Jun 23 2016 14:36
I have my max set to 3, but I haven't hit that yet, as it has been getting limited by MaxSafeBasal
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 14:37
We changed my maxsafebasal
Garrett Webb
@garetis
Jun 23 2016 14:38
I assumed that MaxSafeBasal was variable based on different parameters... e.g., your current BG and assuming it was running for 30 minutes without a new loop being calculated
I don't have my NS set up yet, but I think the idea is that if you see your max basal continually getting used, then you might want to bump it up at that time
Chris Oattes
@cjo20
Jun 23 2016 14:44
I'd say that if you're constantly hitting the max basal, you might need to look at your ratios first
Garrett Webb
@garetis
Jun 23 2016 14:47
...that too.
Sarah Hodson Grady
@shgrady
Jun 23 2016 14:48
@cjo20 sorry, I am still in set up mode on a new pump, on my Omnipod, I think I had it set at 1.5 because I mostly use temp basals for reductions in insulin, square boluses for increases.
Garrett Webb
@garetis
Jun 23 2016 14:50
I'm not sure I follow. Reductions in insulin?
Sarah Hodson Grady
@shgrady
Jun 23 2016 14:53
like, I am going to work out, I want no insulin for the next hour, or I am headed to a low, temp of no insulin until I flatten back out. I was not sure .2 extra an hour would be enough in the new way(openAPS), but it was fine in the old way because I would do square waves when I wanted more insulin.
Garrett Webb
@garetis
Jun 23 2016 14:55
Ah, yeah, gotcha
Sarah Hodson Grady
@shgrady
Jun 23 2016 14:56
unfortunately, I did not understand all the docs when I read at first, so I tend to get ahead of myself since my husband is convinced this project could kill me, and super cautious with trying to understand EVERYTHING, sorry to room, its probably annoying
Garrett Webb
@garetis
Jun 23 2016 15:00
Nah, people like helping out
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 15:04
@garetis I'm not 100% sure but I'm pretty sure there's a hard upper limit on MaxSafeBasal, which is what we changed. I think the default is 3x your current basal
@shgrady I for one am not annoyed one bit! Sharing and asking helps others learn. Certainly it does for me at least.
Garrett Webb
@garetis
Jun 23 2016 15:06
That would be right, because my basal is 0.7 u/hr all day. So, to change your MaxSafeBasal, did you just up your basal rates, to go in a configure that variable?
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 15:06
@Pazoles would have to answer that. He changed something in the code to allow it to go higher than 3x the current.
Paul Martin
@Spazholio
Jun 23 2016 15:06
IIRC, you will never get a temp basal rate higher than your set basal rate * 3. It's hardcoded into OpenAPS.
You'll have to find where that is in the code to change it. I have an idea how to make that more user-configurable, but I need to run it past @scottleibrand @danamlewis and/or @bewest.
Garrett Webb
@garetis
Jun 23 2016 15:07
That's something that wouldn't surprise me
Chris Oattes
@cjo20
Jun 23 2016 15:08
Spazholio: I think you deliberately don't want it easy to configure. It's a pretty good safety feature
Garrett Webb
@garetis
Jun 23 2016 15:10
This brings up a question I was curious about - there's the master code that is written, but do people copy that and up that in their repository and run their own copies? Or does it run the master that is being offered?
Paul Martin
@Spazholio
Jun 23 2016 15:10
@cjo20 Not easily configured, but I think you need to change the "3" to a "4" (or whatever) in multiple locations. I'd rather have constant - something like $MAX_BASAL_SAFETY_COEFFICIENT - that is defaulted to "3" but you can change it to whatever you want in one single location, and then it changes it everywhere. I think I've identified where that needs to be done, just wondering if that is something that would be considered if I were to go ahead with it.
Garrett Webb
@garetis
Jun 23 2016 15:11
I would +1 on that
Chris Oattes
@cjo20
Jun 23 2016 15:11
garetis: it probably depends on the person. I've got some changes in my repository, but I'm planning on commiting those soon
japrobo
@japrobo
Jun 23 2016 15:13
@danamlewis do I need to modify the #223 PR in order to make it ready for you to merge?
Garrett Webb
@garetis
Jun 23 2016 15:23
@cjo20 what would commiting them do? Request them to be merged into the master, or allow your system to run the changes you made?
Chris Oattes
@cjo20
Jun 23 2016 15:24
@garetis I should have said "submit a pull request", which would get them pulled in to the dev branch (if accepted)
Garrett Webb
@garetis
Jun 23 2016 15:25
Is your rig running the changes you created, or are they changes you've written/devised but not implemented into your system?
Chris Oattes
@cjo20
Jun 23 2016 15:26
my rig is running them
Scott Leibrand
@scottleibrand
Jun 23 2016 15:27
We do plan to allow override of the max basal safety coefficient (good name) in preferences.json. I think @sulkaharo was also looking at maybe implementing that. Interested in thoughts on whether we should have another limit on how high that can be overridden without changing code... It's actually 4x current or 3x max_daily_basal right now.
Chris Oattes
@cjo20
Jun 23 2016 15:27
The changes I've written are to make the determine-basal request basal rates the pump can actually deliver (before it just rounded them to the nearest 0.01, which pumps can't do) and display some values in mmol/L in nightscout
Garrett Webb
@garetis
Jun 23 2016 15:31
I like the 0.01 resolution change. So, once any changes get added to the dev branch, would I need to do an upgrade for my version of openaps for my rig to implement those, and until then will just run on the version that I downloaded whenever I started?
Chris Oattes
@cjo20
Jun 23 2016 15:31
@garetis correct. The code I have at the moment is only correct for x23 and x54 devices. It isn't correct for x22
Paul Martin
@Spazholio
Jun 23 2016 15:32
@scottleibrand If you've already got something in the works for preferences.json I won't work on it, but I'm eagerly awaiting what happens. I find myself bumping up against the max rate (5.4u in my case) fairly often.
Chris Oattes
@cjo20
Jun 23 2016 15:32
Spazholio: after meals? Or when fasting?
Paul Martin
@Spazholio
Jun 23 2016 15:34
@cjo20 It's certainly not when fasting. But yeah, after meals, I tend to bolus VERY minimally, and I set temp targets often. I don't eat carbs, so most of my glucose hits over time - usually 2-5 hours after a meal. But if the system sees me ramping up, it tries to mitigate it, but can't do it as aggressively as it wants to because it's hitting that ceiling.
Scott Leibrand
@scottleibrand
Jun 23 2016 15:38
@Spazholio I do not. We just know we need to do it. Check with @sulkaharo to see if he has done anything yet. I think he wanted to refactor the two instances of maxSafeBasal into one function.
Dana Lewis
@danamlewis
Jun 23 2016 15:41
@japrobo wasn’t sure if you’re done. but if you’re ready, I’ll give it a look in a few hours once things calm down at work :)
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 15:49
@Spazholio I wonder if this is a specific low carb issue, because I had the exact same problem after eating. Increasing my max safe helped a lot.
Paul Martin
@Spazholio
Jun 23 2016 15:51
Certainly a possibility. Our glucose load definitely is slower, but it seems to be more...juggernaut-y. =) A spike from carbs can be predicted, but when it's from a low-carb diet, it throws another variable into the mix. And I'm really loathe to switch back to using square wave boluses, simply die to how OpenAPS interprets them currently. =/
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 15:54
I think I've found a sweet (haha) spot with the increased max basal and my two dose system. I never really used square wave before very much so I'm not missing it too much.
But yeah. Protein spikes are weird and often hard to predict for me
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 16:00
All that being said - the original maxsafebasal was still
keeping me in range, just at the upper end rather than the lower. I wasn't going high.
Sarah Hodson Grady
@shgrady
Jun 23 2016 16:11
@Spazholio @bpazoles definitely wondering, I have celiac, and eat mostly vegetables with some meat, starting to swear how fibrous vegetables must be, unique to the plant (I know, I must be the only person who does not hate-hate MONSANTO)
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 16:13
Some veg spikes me much more than others. It's quite variable. I have a really hard time with carrots, onion, tomato, and of course corn.
David Cintron
@loudestnoise
Jun 23 2016 16:23
@scottleibrand So I’m trying out your setup.sh from the advanced-meal-assist branch on your Github and I’m thinking my cron as it is setup now is running every minute. Am I supposed to put in a line of my own that runs openaps loop and the retry checks to see if the loop is running and then if it’s not does retry-loop ?
(pi @ raspberrypi): ~/skadoosh
⇢  crontab -l
SHELL=/bin/bash
PATH=/home/pi/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games:/home/pi/bin

# */5 * * * * (cd /home/pi/openaps && time openaps loop2) 2>&1 | logger -t openaps-loop

* * * * * killall -g --older-than 10m openaps
* * * * * cd /home/pi/skadoosh && oref0-reset-git
* * * * * cd /home/pi/skadoosh && ( ps aux | grep -v grep | grep -q 'openaps retry-loop' || openaps retry-loop ) 2>&1 | tee -a /var/log/openaps/loop.log
Paul Martin
@Spazholio
Jun 23 2016 16:33
@bpazoles What has your dosing strategy been? I'm always open to improvements. :)
japrobo
@japrobo
Jun 23 2016 16:34
@danamlewis not done but eager to get some of the older stuff merged. I'll start a new branch and a new PR and keep going. :) Thanks!
Dana Lewis
@danamlewis
Jun 23 2016 16:36
@japrobo sweet :) good to know!
thanks for all the contributions!
japrobo
@japrobo
Jun 23 2016 16:38
@danamlewis thanks for all your help!
Bill Stackpole
@billstackpole
Jun 23 2016 16:47
@morsen - yes, with enlight setup my cgm and my pump
Are the same.
<grr gitter mobile app and fat fingers>
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 16:50
10 "carbs" a half hour later
Don't know why my first message is highlighted at red - did you see it @Spazholio ?
Bill Stackpole
@billstackpole
Jun 23 2016 16:55
@garetis
Usually you can find the path to any executable in UNIX with the "which" command. E.g. which ls or which bash. You may also want to learn about the help systems in UNIX (aka "man pages" for "manual" pages. Two ways to get there: man <command> and info <command>). Some commands have different man and info page content. Others have only man page content (so the output of info yields contents of man page).
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 16:57
@Spazholio I don't think it went through. @Spazholio I dose for any carbs I eat right at the time of eating them, and then I dose for 40% of the protein a half hour later. So if I ate a meal with 3G carb and 25g protein, I would dose for 3 carbs at the time of eating and 10 "carbs" a half hour later.
Scott Leibrand
@scottleibrand
Jun 23 2016 17:08
@loudestnoise the ps aux | grep -v grep | grep -q 'openaps retry-loop' || openaps retry-loop part checks whether it is running and only starts it if not.
dougtakeuti
@dougtakeuti
Jun 23 2016 17:14
Hey there, I've been away from this channel for a month or so (I hope nobody here ever gets dengue fever...) Is there a place where we keep track of updates? I know that when I became ill I was about to configure autosensitivity and AMA. Any important developments I should be aware as I update my rigs?
Dana Lewis
@danamlewis
Jun 23 2016 17:16
yuck, hope you’re feeling better!
nothing major feature-wise, other than maybe temp targets which require AMA to do so you haven’t missed out
Paul Martin
@Spazholio
Jun 23 2016 17:19
And...my rig's network just died in the middle of lunch. I'm console'd in now, and have no idea what's going on. How odd.
dougtakeuti
@dougtakeuti
Jun 23 2016 17:20
@danamlewis the weirdest tropical disease ever... but I'm ok now, thanks! Ok, thanks for letting me know. What do you mean by temp targets? Is that related to those tweets about activity mode?
Dana Lewis
@danamlewis
Jun 23 2016 17:20
Yes. Or eating soon mode. So you can set it and let the loop take care of it :) (can be entered through NS care portal after you get AMA finished setting up)
dougtakeuti
@dougtakeuti
Jun 23 2016 17:21
sounds handy! :smile:
Paul Martin
@Spazholio
Jun 23 2016 17:28
Ok, so I ended up having some weird entry in my /var/lib/dhcp/dhclient.leases file - something about a Can't allocate interface wlalease error. I have a networking script to make sure I'm always connected, so I'm considering adding a line in there that just purges the /var/lib/dhcp/dhclient.leases file if it can't pull an IP fairly quickly. Is there any downside to doing that, other than I might not get my same IP back?
GabeA23
@GabeA23
Jun 23 2016 18:25

Hey everyone... I'm in Phase 1/2 of my openAPS rig. I'm running into a problem.

I try to invoke my "oref0-monitor/profile.json" report and the terminal output looks fine (no errors). But the "profile.json" file comes out empty. I'm trying to figure out why my "profile.json" file is empty. Any ideas?

Dana Lewis
@danamlewis
Jun 23 2016 18:30
@japrobo I had @scottleibrand do the heavy lifting on that one :)
japrobo
@japrobo
Jun 23 2016 18:33
@danamlewis @scottleibrand yippee!
Scott Leibrand
@scottleibrand
Jun 23 2016 18:35
there is still a lot of stuff I couldn't merge
let us know if you need help cleaning up the merge conflicts. I think in a lot of cases re-applying your changes will be easier, though.
David Cintron
@loudestnoise
Jun 23 2016 18:43
@scottleibrand so does this look right? I noticed it wasn’t uploading to Nightscout and I think that’s because I didn’t have my variables in the cron
(pi @ raspberrypi): ~/skadoosh
⇢  crontab -l
SHELL=/bin/bash
PATH=/home/pi/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/games:/usr/games:/home/pi/bin
NIGHTSCOUT_HOST=http://myhost.nightscout; export NIGHTSCOUT_HOST
API_SECRET=blahblabhblahblahblah; export API_SECRET
*/5 * * * * (cd /home/pi/skadoosh && time openaps loop 2>&1 | tee -a /var/log/openaps/loop.log
* * * * * killall -g --older-than 10m openaps
* * * * * cd /home/pi/skadoosh && oref0-reset-git
* * * * * cd /home/pi/skadoosh && ( ps aux | grep -v grep | grep -q 'openaps retry-loop' || openaps retry-loop ) 2>&1 | tee -a /var/log/openaps/loop.log
Scott Leibrand
@scottleibrand
Jun 23 2016 19:03
you don't need the ; export stuff
David Cintron
@loudestnoise
Jun 23 2016 19:06
ah ok
Ben West
@bewest
Jun 23 2016 19:35
shouldn't need the environment variable at all using the ns device
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 19:55
Any pebble experts here? I'm using an original and lately it's constantly been losing signal. I have to restart the whole thing at least every half hour if not more often, which is pretty annoying. I'm thinking of upgrading to a round but not sure if this will solve the issue. It's been a problem ever since I updated it to the new operating system a few weeks ago.
Sarah Hodson Grady
@shgrady
Jun 23 2016 19:59
also recently having this issue with my round since an AT&T ota update, so I was not blaming the pebble, Marshmellow and the S7 were working better than Kitkat/Note3 had.
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 20:01
I'm using my with an iPhone
Mine*
The watch completely crashed when I updated it and I thought I was going to have to replace it. Matt was able to save it after a LOT of troubleshooting, but ever since then it's been really glitchy.
David Cintron
@loudestnoise
Jun 23 2016 20:12
@bewest Yeah, @scottleibrand’s setup.sh doesn’t use the ns device
I was just playing around with it though. The RileyLink I have stopped working, so went back to the RPi3 rig and thought I’d give this auto-sens thing a try
Sarah Hodson Grady
@shgrady
Jun 23 2016 20:17
@bpazoles you just upgraded to the pebble os with a timeline? if so, might be time, and it is pretty, but I still only have half a camping mode watchface for xdrip.
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 20:18
Yeah the os with the timeline. I really dislike it but I didn't seem to have a choice in not updating it. What's camping mode?
Sarah Hodson Grady
@shgrady
Jun 23 2016 20:21
so simple cgm spark and CGM in the Cloud- Sky allow you to watch Nightscout values (great for parents, but for those of us watching our own blood sugars) sometimes my phone cannot reach the internet, "camping mode" might be the wrong term, but there does not seem to be a watchface for watching xdrip locally on a time round.
somebody sent me one, but some of the parts were on top of each other, then it crashed, now I cannot find it.
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 20:24
Oh, I see. That's good to know.
You do think my connectivity issues would be fixed in general though by upgrading?
I miss not being able to have multiple watch faces on it at the same time. I really don't understand why they got rid of that feature.
Sarah Hodson Grady
@shgrady
Jun 23 2016 20:32
I have a classic my daughter wears, she does not mind because she mostly uses for games. You will get the multi-watchfaces back with the round, but you go into the watchfaces app to switch, I don't mind because I love the timeline which has all my pins, a glucose reading every 5 mins.
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 20:36
Thanks for all the info Sarah!
Sarah Hodson Grady
@shgrady
Jun 23 2016 20:47
@bpazoles no problem, I have like 25 apps on my pebble, its a little nuts.
Lynne Butts
@flutelynne_twitter
Jun 23 2016 20:52
@bpazoles I use the time steel for one of the spark or cgm in the cloud and use it with the share acct. If the numbers are off, I check the pump to see if it is looping. For some reason, the urchin doesn't stay connected for me. My husband's iphone 4 will not allow the pebble time to work on it, so since it only wears it occasionally, he uses one of the android NS rig phones. Works great on that. My classic and his classic did not do well with any of the updates from pebble or ios. Depending where/how you got it, you may be able to exchange it.
Brooke Armstrong Pazoles
@bpazoles
Jun 23 2016 20:54
Oh thanks @flutelynne_twitter, that's all really good info. I got it directly from pebble, maybe I will contact them and see if they will at least give me a discount on an upgrade. They certainly have a record of our customer service queries trying to fix it when the upgrade broke it.
Lynne Butts
@flutelynne_twitter
Jun 23 2016 20:54
You're welcome!
Garrett Webb
@garetis
Jun 23 2016 21:07
Is there anywhere that I can get additional info on the "Boluses" from enact/suggested.json? Mine keeps giving undefined, so not sure if that's an issue or not.
Scott Leibrand
@scottleibrand
Jun 23 2016 21:58
That comes from meal.json.
Jarred Yaw
@jyaw
Jun 23 2016 21:59
@xpucuto any tips on pancreabble? Did a gitter search and saw you were able to get it working... I'm stuck getting a "host is down" error after I've successfully done the first notify. On the 2nd and subsequent notifies I get the error.
Garrett Webb
@garetis
Jun 23 2016 21:59
@scottleibrand I don't recall seeing that anywhere in the docs, is that a part of AMA?
Scott Leibrand
@scottleibrand
Jun 23 2016 21:59
If you have meal assist configured, it'll be the total boluses over the last 1.5* DIA hours or something like that.
Garrett Webb
@garetis
Jun 23 2016 22:00
Gotcha, so it's a part of AMA. Thanks
Scott Leibrand
@scottleibrand
Jun 23 2016 22:00
It compares it to the carbs over the same period to decide if you need a higher assumed basal to cover them. No, it's part of the old meal assist code in master.
AMA actually calculates COB.
Garrett Webb
@garetis
Jun 23 2016 22:01
Ok. How do the carbs get entered, from Bolus Wizard?
Scott Leibrand
@scottleibrand
Jun 23 2016 22:01
Yes, or downloaded from NS.
Sarah Hodson Grady
@shgrady
Jun 23 2016 22:37
blob
still poking through the commands and trying to figure out which create the repoorts I need
what exactly is read_history_data and read_glucose_data supposed to get, pass in nothing, "too few arguments" pass in 1 get "ValueError: need more than 0 values to unpack"
Scott Leibrand
@scottleibrand
Jun 23 2016 22:50
Look at the ones tagged oref0.
Eric
@ecc1
Jun 23 2016 23:28
Just saw that Emutex Labs has orphaned ubilinux for Edison. Probably old news to folks here, but should have seen that coming. I've had partial success transplanting the 3.10.98 kernel from Yocto 3.5 into a ubilinux image, so it might be possible to roll one's own Debian-based system with just the Yocto kernel and modules.
Paul Martin
@Spazholio
Jun 23 2016 23:34
I saw that a month ago, but completely forgot about it. So it looks like Yocto’s it, huh?
Sarah Hodson Grady
@shgrady
Jun 23 2016 23:58
meaning I should only bother with those tagged oref0?
shgrady @shgrady is low again because she super-bolused and got distracted and forgot to enable her temporary basal of zero