Funzler: a Functional Deficiency Diagnosis Tool

Funzler is a prototype tool for functional deficiency diagnosis. This is generated during the PhD work of the author.

Dependencies

1. Required python libraries include numpy, pandas, pyyaml.
2. Knowledge base should be available in the Causal Scenario Analysis (CSA). For the execution of a CSA, a template is provided. More detailed methodological aspects can be found in the dissertation.

Usage in Terminal

1. Configure the knowledge base (from CSA) in funzler.yaml.

   • As an example of the CSA, CSA_Example.xlsx is provided. The CSA results have been generated during Experiment 2 of the dissertation.

2. Start and run a funzler session (a diagnosis job).

   • The script will inquire user for the observation of trigger-events. For each trigger-event, a value from {1, -1, 0} should be given, respectively representing {"present", "absent", "unknown"};
   • For example, for a 5-events-observation, an user input 1 1 1 1 -1 means the 5th trigger-event is absent and others are present.

   python3 funzler_start.py
   

3. If the previous script outputs a label "fail pending" and some measurement updates are possible, e.g. via human annotation, update and rerun the funzler session (diagnosis job).

   • The script will inquire user for the updated boundary states, which are seperated by spaces
   • For example, 1 1 23 1 34 0 means the boundary id 1 is updated as present (or intensity index equal 1.0), the boundary id 23 as present, and the boundary id 34 as absent.

   python3 funzler_update.py
   

Usage in Jupyter Notebook

Configure the knowledge base in funzler.yaml and run in terminal:

jupyter notebook funzler.ipynb