Skip to content

My personal branch of an adaptation of some parts of AAPS

License

Notifications You must be signed in to change notification settings

jgslade/AAPS-EatingNow

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AndroidAPS

Support Server

Build status Crowdin Documentation Status codecov dev: codecov

BTC 3KawK8aQe48478s6fxJ8Ms6VTWkwjgr9f2

######################################## This version of AAPS has evolved over time using elements from AIMI and Boost. This AAPS variation is called "Eating Now" (EN) as it is a reactive operating mode without needing to inform the system. The intent of this plugin is the same, to deliver insulin earlier using mostly openAPS predictions. This has been tested successfully with a blend of Fiasp 80% and Novorapid 20% (F80N). The code can be used with or without bolusing or COB entries. However it will not become active until a treatment has been performed after the active start time. Allowing a delayed start time for example if you sleep in. :) This treatment can be 1g or 0.1U for example. After this the EN mode is active until the end time specified in the preferences. It is recommended to set maxSMBBasalMinutes and maxUAMSMBBasalMinutes to 60 minutes max as these will be used when EN is not active.

These are the methods utilised within this version:

UAM: This is based upon Boost and is used when there is a sudden increase in BGL of >=9 (0.5 mmol) UAMBoost will only operate when there are no COB. TDD is used as a reference point for initial insulin dose that can be scaled within preferences.

COB: When carbs are entered there is a time window like AIMI. The COBpredBG prediction uses the dynamic ISF from Boost to increase insulinReq. If within the COBBoost Window the calculated insulinReq may be delivered via a larger SMB using the COBBoost maxBolus. Once the time window has elapsed COBBoost maxBolus is no longer used.

Predictions leverage the dynamic ISF concept within the Boost plugin. Using the eventualBG mostly to determine the insulinReq. The main difference is the initial ISF used to determine the predictions is based on the profile ISF. If BG is currently the normalTarget BG from the profile the ISF will be the same as the profile. Once BG rises the ISF number reduces, and as BG lowers the ISF number will increase. ISF scaling can be adjusted and eventualBG weighting for UAM and COB predictions can be applied.

These are the preferences utilised for EN mode:

General: Start Time: The time the EN mode will start in hours as 24h clock format EN mode will be active after this time when there has been a COB or manual bolus entry of any size End Time: The time that EN mode will finish. Normal maxBolus of 65% is resumed. If there are COB or a TT of normalTarget EN will be active after this time, however AAPS maxBolus will be used. No SMB will be given when inactive unless there is COB, detected resistance from autosens or BG is above SMB BG Threshold. InsulinReqPct: Percentage that will be used for EN insulinReq as SMB to utilise prior to maxBolus restriction. This will be 65% when EN is not active. Max IOB: The percentage of current max-iob setting that will be used as the limit for EN. EN will not add insulin when above this limit. SMB BG Offset: There will be no SMB when below this BG offset at night without COB or detected resistance. e.g. if target bg is 99/5.5 and this setting is 27/1.5 there will be no SMB below 126/7.0 ISF BG Scaler: As BG increases ISF will become stronger. The level of scaling can be adjusted. 0 = normal scaling, 5 is 5% stronger, -5 is 5% weaker ISF scaling. Additional scaling does not happen when EN is not active. ISF BG Offset: As BG increases ISF will become stronger. ISF will no longer scale when above this level. e.g. if target bg is 99/5.5 and this setting is 27/1.5 there will be no scaling above 126/7.0

UAM: UAMBoost Bolus Scale: Multiply the initial UAMBoost bolus by this amount. 0 will disable UAMBoost. UAM maxBolus: maxBolus to use for all BG rises without COB. 0 will use maxSMBBasalMinutes or maxUAMSMBBasalMinutes.

COB: Use GhostCOB: Ignore COB predictions after the COBBoost Window and rely purely on UAM. This setting can be handy when COB lingers for too long. COBBoost InsulinReqPct: Percentage that will be used for EN insulinReq within the COBBoost Window. COBBoost Window: If within the COBBoost Window the calculated insulinReq from COBPredBG may be delivered via a larger SMB using the COBBoost maxBolus. Once the time window has elapsed COBBoost maxBolus is no longer used. 0 minutes will disable this functionality. COBBoost maxBolus: maxBolus to use within the COBBoost Window. 0 will use AAPS maxBolus. COB maxBolus: maxBolus to use with COB outside of the initial COBBoost Window. 0 will use AAPS maxBolus.

EXPERIMENTAL: • Use 3PM Basal ISF Variance: Use 3PM Basal as the basis for ISF changes. The basal at 3PM is taken as the basis for baseline ISF. Basal variation from this point is used to scale the ISF, stronger basal will make ISF weaker. Only use when the profile uses a single ISF for 24 hours and the basal profile is fully populated. • Use TDD for ISF: Use the last 24H TDD for ISF. This will override the profile ISF and can be used with 3PM basal ISF variance. • TDD ISF Scaling: This will use a percentage of the calculated TDD ISF. If TDD ISF is too strong it can be reduced e.g. 50 will make TDD ISF 50% weaker.

About

My personal branch of an adaptation of some parts of AAPS

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Kotlin 78.3%
  • Java 17.4%
  • JavaScript 4.3%