-
Notifications
You must be signed in to change notification settings - Fork 0
/
3_feature_scaling.py
35 lines (26 loc) · 1.26 KB
/
3_feature_scaling.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
from sklearn.preprocessing import StandardScaler
import pandas as pd
import os
# Load the extracted features from the Excel file
df = pd.read_excel("D:/MCA/4th sem/SER3/output/Actor_01/output_data.xlsx")
# Separate the MFCC features
mfcc_columns = [col for col in df.columns if col.startswith('MFCC')]
X_mfcc = df[mfcc_columns]
# Initialize the StandardScaler
scaler = StandardScaler()
# Fit the scaler to the MFCC features and transform them
X_mfcc_scaled = scaler.fit_transform(X_mfcc)
# Convert the scaled MFCC features back to a DataFrame
scaled_df_mfcc = pd.DataFrame(X_mfcc_scaled, columns=mfcc_columns)
# Add the 'Audio File' and 'Emotion' columns back to the DataFrame
scaled_df_mfcc['Audio File'] = df['Audio File']
scaled_df_mfcc['Emotion'] = df['Emotion']
# Reorder the columns to have 'Audio File' first, followed by MFCC and Emotion
column_order = ['Audio File'] + mfcc_columns + ['Emotion']
scaled_df_mfcc = scaled_df_mfcc[column_order]
# Display the scaled MFCC DataFrame
print(scaled_df_mfcc.head())
# Save the scaled MFCC features to a new Excel file in the output folder
output_folder = "D:/MCA/4th sem/SER3/output/Actor_01"
scaled_mfcc_excel_file = os.path.join(output_folder, "scaled_output_data.xlsx")
scaled_df_mfcc.to_excel(scaled_mfcc_excel_file, index=False)