search

Pi-Star-Lab / RESCO

Reinforcement Learning Benchmarks for Traffic Signal Control (RESCO)
107 stars 36 forks source link

Result interpretation #26

Open Kunal-Kumar-Sahoo opened 4 weeks ago

Kunal-Kumar-Sahoo commented 4 weeks ago

Some metrics_i.csv file is being generated for every $i^{th}$ route while training any model. Now each observation record (every $10^{th}$ second) has three dictionaries containing numerical values against each key (which are the junction IDs).

How do I interpret those results? I want to specifically know what these three dictionaries represent?

jault commented 4 weeks ago

Yea, these ought to have been saved with headers. You can find the definition over here: https://github.com/Pi-Star-Lab/RESCO/blob/3c51ea43764a10dc3b6eb3fcf2e8bf503eac1cf1/resco_benchmark/multi_signal.py#L219