How to use the experiment to make the figures?
w1w9 opened this issue · 4 comments
Hello sir, it is my honor to study your project. I can understand the rough process, but I do not know how to use the experiment results and use them to make the figures like that. Could you give me an instruction about the experimental indicators ?(logs, placements in script.py and ray results in tuning.py). I am looking forward to your reply very much.
Hello sir, it is my honor to study your project. I can understand the rough process, but I do not know how to use the experiment results and use them to make the figures like that. Could you give me an instruction about the experimental indicators ?(logs, placements in script.py and ray results in tuning.py). I am looking forward to your reply very much.
Hello, could you share this code with me for learning? Thank you!
I'm happy you are working with our project! Sorry for the late reponse.
@stwerner97 @NilsLuca Does one of you remember in which format the results were stored and could share some tips about how to plot them?
Yes, sure, sorry for the late response!
Looking at the EvalLogCallback, we store statistics about Eval. Iter
, Episode
, Arrival
, TTL
, Bandwidth
, Max Latency
, VNFs
, and Placements
at each step(?) of the simulation. The VNFs
field is a list that denotes the components of the SFC
and their resource requirements (cpu & memory).
In terms of plotting our results, I just collected the statistics of each episode in a pandas dataframe, aggregated the data (e.g., computed the average total reward of episodes) and visualized them using seaborn & matplotlib.