ATM Repository

Repository containing code, as well as gathered data, as used for the paper

Merlijn Krale, Thiago D. Simao, Jana Tumova, Nils Jansen
Robust Active Measuring under Model Uncertainty
In AAAI, 2024.

Contents

This repository contains the following files:

Code:

• ACNO_Planning.py : Code containing all planning algorithms used in the paper;
  • Note: Measurement lenient algorithsm are refered to as 'Control-Robust'.
• Run.py : Code for automatically running agents on environments & recording their data;
• RunAll.sh : Bash file for automatically running all experiments in the paper;
• Plot_Data.ipynb : Code for plotting data (with a matplotlibrc file to set formatting);
• Requirements.text : File with required python dependencies;

Folders:

• AM_Gyms : Contains all code related to setting up and learning models, as used by the planning algorithms.
• Data : Contains gathered data, including analysed data & plots.
• Baselines : Contains code for all baseline algorithms used while testing.

Getting started

After cloning this repository:

1. create a virtualenv and activate it

cd ATM/
python3 -m venv .venv
source .venv/bin/activate

2. install the dependencies

pip install -r requirements.txt

How to run

All algorithms can be run using the Run.py file from command line. Running 'python Run.py -h' gives an overview of the functionaliality.

As an example, starting a run looks something like:

python Run.py -alg ATM_Control_Robust -env Drone -alpha_plan 0.5 -alpha_real 0.8 -alpha_measure 0.8 -nmbr_eps 100

This command runs the MLATM algorithm on the Drone environment with $\alpha = 1, \alpha_p = 0.5$, and $\mathcal{M}_\text{ML}$ with dynamics parametrized an RMDP with $\alpha=0.8$. Thus, CR-ATM-avg uses alpha_measure 1, CR-ATM-pes uses alpha_measure = alpha_plan, and CR-ATM-opt uses alhpa_measure = - alpha_plan (hard-coded). To run all experiments from the paper at once, run the following:

bash ./Runall.sh

Note that this repository does not contain all pre-computated files for the drone environments, which means run times might be relatively long.