No pandas

tests/python/physics/lgrngn_cond_substepping.py

@@ -21,7 +21,7 @@
 from libcloudphxx import lgrngn
 from libcloudphxx import common
 
-import pandas as pd
+import csv
 
 # wrapper for timing excecution time
 def wrapper(func, opts, th, rv, rhod):
@@ -284,13 +284,29 @@ def test(RH_formula, step_count, substep_count, exact_substep, constp, mixing, a
               results['sstp_cond_adapt_drw2_max'] = opts_init.sstp_cond_adapt_drw2_max
               records.append(results)
 
+def _write_csv(path, rows):
+  os.makedirs(os.path.dirname(path), exist_ok=True)
+  # Collect fieldnames deterministically (preserve insertion order from first row, then add any missing keys)
+  fieldnames = []
+  seen = set()
+  for r in rows:
+    for k in r.keys():
+      if k not in seen:
+        fieldnames.append(k)
+        seen.add(k)
+  with open(path, "w", newline="") as f:
+    w = csv.DictWriter(f, fieldnames=fieldnames)
+    w.writeheader()
+    for r in rows:
+      w.writerow(r)
+
 # save results to a CSV file for refdata comparison and for plotting
-df = pd.DataFrame(records)
-df['sd_conc'] = opts_init.sd_conc  # Add the column to all rows at once
-df['RH_max'] = opts_init.RH_max
-df['dt'] = 1
-os.makedirs("test_results", exist_ok=True)
-df.to_csv("test_results/lgrngn_cond_substepping_results.csv", index=False)
+for r in records:
+  r['sd_conc'] = opts_init.sd_conc
+  r['RH_max'] = opts_init.RH_max
+  r['dt'] = 1
+
+_write_csv("test_results/lgrngn_cond_substepping_results.csv", records)
 
 # Optionally save as reference data
 if '--save-ref' in sys.argv:
@@ -300,8 +316,8 @@ def test(RH_formula, step_count, substep_count, exact_substep, constp, mixing, a
   refdata_dir = os.path.join(script_dir, "refdata")
   os.makedirs(refdata_dir, exist_ok=True)
   refdata_file = os.path.join(refdata_dir, "lgrngn_cond_substepping_refdata.csv")
-  df.to_csv(refdata_file, index=False)
-  print("Reference data saved to: test_results/lgrngn_cond_substepping_refdata.csv")
+  _write_csv(refdata_file, records)
+  print(f"Reference data saved to: {refdata_file}")
   print("Future runs can be compared against this reference using:")
   print("  python lgrngn_cond_substepping_test.py")
 

tests/python/physics/lgrngn_cond_substepping_plot.py

@@ -1,9 +1,32 @@
-import pandas as pd
+import csv
 import matplotlib.pyplot as plt
 import os
 
+
+def _read_csv_rows(path):
+    with open(path, newline="") as f:
+        return list(csv.DictReader(f))
+
+
+def _to_bool(v):
+    s = str(v).strip().lower()
+    if s in ("true", "1", "t", "yes", "y"):
+        return True
+    if s in ("false", "0", "f", "no", "n"):
+        return False
+    raise ValueError(f"Cannot parse boolean from: {v!r}")
+
+
+def _to_float(v):
+    return float(str(v).strip())
+
+
+def _to_int(v):
+    return int(float(str(v).strip()))
+
+
 # Load the results
-df = pd.read_csv("test_results/lgrngn_cond_substepping_results.csv")
+rows = _read_csv_rows("test_results/lgrngn_cond_substepping_results.csv")
 
 variables = [
     ('ss', 'Supersaturation [%]'),
@@ -16,25 +39,48 @@
     ('exectime', 'Execution time [s]'),
 ]
 
-RH_formula_names = {
-    0 : "pv_cc", 1: "rv_cc", 2 :"pv_tet", 3 : "rv_tet"
-}
+# RH_formula_names = {
+#     0 : "pv_cc", 1: "rv_cc", 2 :"pv_tet", 3 : "rv_tet"
+# }
 
 group_cols = ['mixing', 'constp', 'exact_sstp', 'adaptive', 'sstp_cond_act']
-RH_formulas = sorted(df['RH_formula'].unique())
+# RH_formulas = sorted({ _to_int(r['RH_formula']) for r in rows })
+RH_formulas = sorted({ r['RH_formula'] for r in rows })
 
 os.makedirs("test_results/plots", exist_ok=True)
 
+def _group_key(row):
+    return (
+        _to_bool(row['mixing']),
+        _to_bool(row['constp']),
+        _to_bool(row['exact_sstp']),
+        _to_bool(row['adaptive']),
+        _to_int(row['sstp_cond_act']),
+    )
+
+
 for var, ylabel in variables:
     fig, axs = plt.subplots(2, 2, figsize=(14, 10), sharex=True)
     axs = axs.flatten()
     for i, RH_formula in enumerate(RH_formulas):
         ax = axs[i]
-        subdf = df[df['RH_formula'] == RH_formula]
-        for name, group in subdf.groupby(group_cols):
-            label = ', '.join(f"{col}={val}" for col, val in zip(group_cols, name))
-            ax.plot(group['sstp_cond'], group[var], marker='o', label=label, linestyle='--')
-        ax.set_title(f'RH_formula={RH_formula_names[RH_formula]}')
+        subrows = [r for r in rows if r['RH_formula'] == RH_formula]
+
+        grouped = {}
+        for r in subrows:
+            k = _group_key(r)
+            grouped.setdefault(k, []).append(r)
+
+        for key, group in grouped.items():
+            label = ', '.join(
+                f"{col}={val}" for col, val in zip(group_cols, key)
+            )
+            group_sorted = sorted(group, key=lambda r: _to_int(r['sstp_cond']))
+            x = [_to_int(r['sstp_cond']) for r in group_sorted]
+            y = [_to_float(r[var]) for r in group_sorted]
+            ax.plot(x, y, marker='o', label=label, linestyle='--')
+        # ax.set_title(f'RH_formula={RH_formula_names[RH_formula]}')
+        ax.set_title(f'RH_formula={RH_formula}')
         ax.set_xlabel('Number of substeps')
         ax.set_ylabel(ylabel)
         ax.set_xscale('log')