Added argps to visualizer to handle different max_evals in single

metrics file

scripts/analysis/visualize_results.py

@@ -1,3 +1,4 @@
+import argparse
 import ast
 import os
 
@@ -6,7 +7,7 @@
 from nmrcraft.analysis.plotting import plot_bar, plot_metric
 
 
-def load_results(results_dir: str, baselines_dir: str):
+def load_results(results_dir: str, baselines_dir: str, max_evals: int):
     import_filename_base = os.path.join(baselines_dir, "results_baselines.csv")
     import_filename_one = os.path.join(results_dir, "results_one_target.csv")
     import_filename_multi = os.path.join(
@@ -15,8 +16,9 @@ def load_results(results_dir: str, baselines_dir: str):
 
     df_base = pd.read_csv(import_filename_base)
     df_one = pd.read_csv(import_filename_one)
+    df_one = df_one[df_one["max_evals"] == max_evals]
     df_multi = pd.read_csv(import_filename_multi)
-
+    df_multi = df_multi[df_multi["max_evals"] == max_evals]
     return df_base, df_one, df_multi
 
 
@@ -134,13 +136,29 @@ def plot_exp_3(df_one, df_multi):
     return
 
 
+# Setup parser
+parser = argparse.ArgumentParser(
+    description="Train a model with MLflow tracking."
+)
+
+parser.add_argument(
+    "--max_evals",
+    type=int,
+    default=100,
+    help="How many max_evals the analysed data has",
+)
+# Add arguments
+args = parser.parse_args()
+
 if __name__ == "__main__":
 
     if not os.path.exists("./plots/results/"):
         os.makedirs("./plots/results/")
 
     df_base, df_one, df_multi = load_results(
-        results_dir="metrics/20eval/", baselines_dir="metrics/"
+        results_dir="metrics/20eval/",
+        baselines_dir="metrics/",
+        max_evals=args.max_evals,
     )
 
     plot_exp_1(df_base, df_one)