Flatten MDP problem using parameters in the problems/rescue.yaml file into temp/problem.h5:

python src/flatten.py -i problems/rescue.yaml -o temp/problem.h5

Solve the flat MDP problem temp/problem.h5 with discount rate 0.9, and store policy to file temp/policy.h5:

python src/solve.py -i temp/problem.h5 -o temp/policy.h5 -d 0.9

Parse the policy file temp/policy.h5 into a high-level policy file sols/rescue.yaml:

python src/parse-policy.py -i problems/rescue.yaml -s temp/policy.h5 -o sols/rescue.yaml

Plot the results of the policy file sols/rescue.yaml:

python src/plot.py -i problems/rescue.yaml -s sols/rescue.yaml -o sim/fig/state_%09d.png

Simulate the policy file sols/rescue.yaml over 10 steps and generate graphic HTML report inside directory sim/:

python src/simulate.py -i problems/rescue.yaml -s sols/rescue.yaml -x 10 -d sim/ -f fig/state_%09d.png -o run.html -g anim.gif -a 150

Solving problem problems/rescue.yaml using Planning Alone and storing the high-level policy sols/rescue.yaml:

./scripts/pa.sh problems/rescue.yaml sols/rescue.yaml

Solving problem problems/rescue.yaml using Coalition Formation and Planning and storing the high-level policy sols/rescue.yaml:

./scripts/cfp.sh problems/rescue.yaml sols/rescue.yaml