style: format

carps/analysis/gather_data.py

@@ -285,6 +285,7 @@ def convert_mixed_types_to_str(logs: pd.DataFrame, logger=None) -> pd.DataFrame:
         logs[c] = logs[c].astype("str")
     return logs
 
+
 def load_set(paths: list[str], set_id: str = "unknown") -> tuple[pd.DataFrame, pd.DataFrame]:
     logs = []
     for p in paths:

carps/analysis/plot_relative_perf.py

@@ -1,46 +1,39 @@
 from __future__ import annotations
 
 import matplotlib.pyplot as plt
-import seaborn as sns
 import pandas as pd
-
-from carps.analysis.utils import savefig
-from carps.analysis.gather_data import process_logs
+import seaborn as sns
 
 
 def norm_by_opt(df: pd.DataFrame, optimizer_id: str) -> pd.DataFrame:
     df_new = []
-    for gid, gdf in df.groupby(by=["problem_id", "seed"]):
-        reference = gdf[gdf["optimizer_id"]==optimizer_id]["trial_value__cost"]
-        gdf["trial_value__cost_normopt"] = gdf.groupby("optimizer_id")["trial_value__cost"].transform(lambda x: x / reference)
+    for _gid, gdf in df.groupby(by=["problem_id", "seed"]):
+        reference = gdf[gdf["optimizer_id"] == optimizer_id]["trial_value__cost"]
+        gdf["trial_value__cost_normopt"] = gdf.groupby("optimizer_id")["trial_value__cost"].transform(
+            lambda x: x / reference
+        )
         df_new.append(gdf)
     df = pd.concat(df_new).reset_index(drop=True)
 
-
     # df["trial_value_cost_normopt"] = df.groupby("problem_id").apply(_norm_by_opt)
-    df["trial_value__cost_inc_normopt"] = df.groupby(by=["problem_id", "optimizer_id", "seed"])["trial_value__cost_normopt"].transform("cummin")
+    df["trial_value__cost_inc_normopt"] = df.groupby(by=["problem_id", "optimizer_id", "seed"])[
+        "trial_value__cost_normopt"
+    ].transform("cummin")
     return df
-
-
-
 
 
-
-if __name__=="__main__":
-    df = pd.read_parquet("/scratch/hpc-prf-intexml/cbenjamins/repos/CARP-S-Experiments/lib/CARP-S/runs/RandomSearch/MFPBench/logs.parquet")
+if __name__ == "__main__":
+    df = pd.read_parquet(
+        "/scratch/hpc-prf-intexml/cbenjamins/repos/CARP-S-Experiments/lib/CARP-S/runs/RandomSearch/MFPBench/logs.parquet"
+    )
 
     df_new = df.copy()
     df_new["optimizer_id"] = "HeheOpt"
     df_new["trial_value__cost"] -= 0.1
     df = pd.concat([df, df_new]).reset_index(drop=True)
 
-    fig, ax = plt.subplots(figsize=(6,4))
+    fig, ax = plt.subplots(figsize=(6, 4))
     normalize_by_opt = "RandomSearch"
     df = norm_by_opt(df, "RandomSearch")
 
     sns.lineplot()
-
-
-
-
-