Update , explore and visualize the data within

.gitignore

@@ -127,3 +127,12 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
+.ipynb_checkpoints 
+<<<<<<< HEAD
+.vscode
+.xlsx
+=======
+
+.vscode
+.xlsx
+>>>>>>> exploratory-ananlysis

.vscode/settings.json

@@ -0,0 +1,3 @@
+{
+    "vscode-pets.throwBallWithMouse": true
+}

exploratory_analysis.py

@@ -0,0 +1,152 @@
+import os
+import wfdb
+import numpy as np
+import pandas as pd
+import neurokit2 as nk
+from pathlib import Path
+from collections import Counter
+import itertools
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+def categorize_records(folder_path):
+    """ 
+    Categorize record files with their labels.
+
+    Parameters:
+    - folder_path (str): The path to the folder containing WFDB records.
+
+    Returns:
+    - records_category (dict): A dictionary mapping file names to their corresponding labels.
+    """
+    labels = []
+    file_names = []
+    files = os.listdir(folder_path)
+    files = [f[:-4] for f in files if f.endswith(".hea")]
+
+    records_category = []    
+    for file in files:
+        try:
+            label = wfdb.rdrecord(os.path.join(folder_path, file)).comments
+            record_info = {'file_name': file, 'label': label}
+            records_category.append(record_info)
+        except Exception as e:
+            print(f"Error processing {file}: {e}")
+
+    return records_category
+
+def get_duration(folder_path):
+    """
+    Calculate the duration of each record in the specified folder_path.
+
+    Parameters:
+    - folder_path (str): The path to the folder containing WFDB records.
+
+    Returns:
+    - duration_list (list): A list of durations corresponding to each record.
+    """
+    records_duration = []
+    files = os.listdir(folder_path)
+    files = [f[:-4] for f in files if f.endswith(".hea")]
+
+    for file in files:
+        record_path = os.path.join(folder_path, file)
+        try:
+            record_data = wfdb.rdsamp(record_path)
+            sample_point = record_data[0]
+            sampling_frequency = record_data[1]['fs']
+            duration = len(sample_point) / sampling_frequency
+            record_info = {'file_name': file, 'duration': duration}
+            records_duration.append(record_info)
+        except Exception as e:
+            print(f"Error processing {record_path}: {e}")
+
+    return records_duration
+
+def get_annotation(folder_path):
+
+    """
+    Extract annotations for record files in the specified folder_path.
+
+    Parameters:
+    - folder_path (str): The path to the folder containing WFDB annotation files.
+
+    Returns:
+    - annotations_category (dict): A dictionary mapping file names to their corresponding annotations and their summary.
+    """
+
+    records_annotations = []
+    annotations_summary = []
+    file_names = []
+    files = os.listdir(folder_path)
+    files = [f[:-4] for f in files if f.endswith(".atr")]
+
+    for file in files:
+        file_names.append(file)
+        record_path = os.path.join(folder_path, file)
+        try:
+            record_annotation = wfdb.rdann(record_path, "atr")
+
+            # Extract annotation information
+            annotation_symbols = record_annotation.symbol
+            annotation_sample = record_annotation.sample            
+
+            # Example: Create a summary using Counter
+            summary = Counter(annotation_symbols)
+
+            # Append relevant information to the lists
+            records_annotations.append({
+                'symbols': annotation_symbols,
+                'sample': annotation_sample                
+            })
+
+            annotations_summary.append(summary)
+        except Exception as e:
+            print(f"Error processing {record_path}: {e}")
+
+    # Pair each file with its corresponding annotations and summary
+    annotations_category = {file_name: {'annotations': ann, 'summary': summ}
+                            for file_name, ann, summ in zip(file_names, records_annotations, annotations_summary)}
+
+    return annotations_category
+
+def create_records_dataframe(folder_path):
+    # Load data into dictionaries
+    categorized_records = categorize_records(folder_path)
+    duration_data = get_duration(folder_path)
+    annotation_data = get_annotation(folder_path)
+    # Create a DataFrame
+    df = pd.DataFrame({'folder_name':os.path.basename(folder_path),
+                        'file_name': [record['file_name'] for record in categorized_records],
+                        'labels': [record['label'][0] for record in categorized_records],
+                        'duration(s)': [record['duration'] for record in duration_data],
+                        'annotation': [annotation_data[file]['annotations'] for file in [record['file_name'] for record in categorized_records]],
+                        'summary_annotation': [annotation_data[file]['summary'] for file in [record['file_name'] for record in categorized_records]]
+                    })
+    return df
+
+def visualize_dataset(df):
+    plt.hist(df['duration(s)'], bins=10, color='blue', alpha=0.7)
+    plt.title('Distribution of Record Durations')
+    plt.xlabel('Duration (seconds)')
+    plt.ylabel('Frequency')
+    plt.show()
+
+
+    plt.figure(figsize=(10, 6))
+    sns.countplot(x='labels', data=df, palette='viridis')
+    plt.title('Distribution of Labels of ')
+    plt.xlabel('Labels')
+    plt.ylabel('Count')
+    plt.xticks(rotation=45, ha='right')  # Rotate x-axis labels for better visibility
+    plt.show()   
+    return 
+
+TSOne_df=create_records_dataframe("D:/RA/Project/Database/CPSC2021/Training_set_I")
+#TSOne_df.to_excel("CPSC2021_TS1.xlsx",index=False)
+visualize_dataset(TSOne_df)
+
+TSTwo_df=create_records_dataframe("D:/RA/Project/Database/CPSC2021/Training_set_II")
+#TSOne_df.to_excel("CPSC2021_TS1.xlsx",index=False)
+visualize_dataset(TSTwo_df) 
+

requirements.txt

@@ -1,7 +1,10 @@
-numpy==1.16.2
-scipy==1.2.1
-wfdb==2.2.1
-matplotlib==3.0.3
-pandas==0.24.2
-peakutils==1.3.3
-scikit_learn==0.24.1
+numpy
+scipy
+wfdb
+matplotlib
+pandas
+peakutils
+scikit_learn
+neurokit2
+tensorflow
+