danamlewis / OpenAPS

The prediction algorithms for OpenAPS

OpenAPS

This repository holds the prediction algorithms for OpenAPS.

The OpenAPS data must be in a subdirectory called "data" with a subdirectory of the ID number that contains the devicestatus.json file. For example:

    ./data/00897741/devicestatus.json
    ./data/01352464/devicestatus.json
    ./data/01884126/devicestatus.json
    ./data/14092221/devicestatus.json
    ./data/15634563/devicestatus.json
    ./data/17161370/devicestatus.json
    ./data/24587372/devicestatus.json
    ./data/40997757/devicestatus.json
    ./data/41663654/devicestatus.json
    ./data/45025419/devicestatus.json
    ./data/46966807/devicestatus.json
    ./data/68267781/devicestatus.json
    ./data/84984656/devicestatus.json

where . represents the current directory with the code

The code requires the following files:

    bgdataframe.py
    bgarray.py
    datamatrix.py
    mlalgorithm.py
    ClarkeErrorGrid.py

This code also requires the following libraries:

    pandas
    numpy
    gatspy
    sklearn

Once all of these have been satsified, you can run the code with the following command:

    python main.py

Trevor Tsue

2017-7-24