combine.py

import pandas as pd
from pathlib import Path

# 1. Define your parent directory path
parent_dir = Path(r"C:\Users\Richard\_Vork\MVT\DONE")

# 2. This list will hold individual dataframes
df_list = []

# 3. Iterate through each item in the parent directory
for group_folder in parent_dir.iterdir():
    if group_folder.is_dir():
        # rglob looks for the filename recursively within the group folder
        # It returns a generator of all matches
        files = list(group_folder.rglob("all_runs_summary_combined_averaged.csv"))
        
        if files:
            # We take the first match found in this specific group folder
            file_path = files[0]
            print(f"Found file in {group_folder.name}: {file_path}")
            
            # Read the CSV and optionally add a column to identify the group
            temp_df = pd.read_csv(file_path)
            temp_df['group_source'] = group_folder.name
            
            df_list.append(temp_df)
        else:
            print(f"No file found in {group_folder.name}")

# 4. Concatenate all found dataframes into one
if df_list:
    combined_df = pd.concat(df_list, ignore_index=True)
    print("\nSuccessfully combined all files!")
    print(f"Total rows: {len(combined_df)}")
else:
    print("No files were found to combine.")

# Show the first few rows
print(combined_df.head())

##############################################

import matplotlib.pyplot as plt
import seaborn as sns

# 1. Set the visual style
sns.set_theme(style="whitegrid")

# 2. Sort the dataframe by group to ensure consistent ordering on the X-axis
combined_df = combined_df.sort_values('group')

# 3. Create the Bar Plot
# 'errorbar=None' ensures only the mean bar is shown without standard error bars
ax = sns.barplot(
    data=combined_df, 
    x='group', 
    # y='void', 
    # y='leak', 
    y='weighted_avg_void_vol', 
    color='skyblue', 
    alpha=0.5, 
    errorbar=None
)

# 4. Overlay the Individual Points
# 'jitter=True' prevents points from overlapping perfectly so you can see density
sns.stripplot(
    data=combined_df, 
    x='group', 
    # y='void', 
    # y='leak', 
    y='weighted_avg_void_vol', 
    color='black', 
    jitter=0.2, 
    size=5,
    alpha=0.7
)

# 5. Labeling and Formatting
# plt.title('Void Count per Group')
# plt.title('Leak Count per Group')
plt.title('AVV per Group')
plt.xlabel('Group Name')
# plt.ylabel('Void Count')
# plt.ylabel('Leak Count')
plt.ylabel('AVV (ul)')

# Tight layout ensures labels aren't cut off
plt.show()
# plt.savefig('void_count_plot.png', bbox_inches='tight')
print("Plot saved as void_count_plot.png")