Team Michal Chromcak

[MichalChromcak]

Team Members

• Michal Chromcak

Your Approach

Just adding a possibility to control majority of constants

• to make it easier - shared functionality for all agents
• allowing sum of both days of inbound_shipments
• deducting next_incoming_order from current_stock to compute base for order_to_place
• setup little FW to store results of different experiments (not attached notebook), that runs following loop

    for experiment in experiments:
          total_costs = {}
          total_costs["model"] = experiment
          total_costs["commit_hash"] = commit_hash
          total_costs["results"] = []
          for i in range(500):
              total_costs["results"].append(run(verbose_sim=False, params=total_costs["model"]))
          store_results(total_costs, csv_path=csv_path)

• and evaluates the results looking at the violin plot with quartiles

  ax = sns.violinplot(x="results", y="model", data=df_best, palette="light:g", inner="points", orient="h")
  sns.violinplot(x="results", y="model", data=df_best, palette="light:g", inner="quartiles", orient="h", ax=ax)
  

  

[github-actions[bot]]

Awesome job, Team Michal Chromcak! Your result has been published to the leaderboard

Your result and this comment will be updated automatically whenever you push new changes. Please note that we'll always keep your last result, not necessarily your best result.

Last submission at 2021-10-30T13:51:19 with costs 258.9.