4/22/25, 1:36 AM Untitled4.

ipynb - Colab

# BigMart Sales Analysis Project

# Predicting sales of products in BigMart stores

# Import necessary libraries

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder, StandardScaler
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error
import warnings
warnings.filterwarnings('ignore')

# BigMart Sales Analysis Project

# Predicting sales of products in BigMart stores

# Import necessary libraries

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder, StandardScaler
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error
import warnings
warnings.filterwarnings('ignore')

# Import data using kagglehub

try:
# Install dependencies if needed
# pip install kagglehub[pandas-datasets]
import kagglehub
from kagglehub import KaggleDatasetAdapter

# Load train data

print("Loading train data from Kaggle...")
train = kagglehub.load_dataset(
KaggleDatasetAdapter.PANDAS,
"brijbhushannanda1979/bigmart-sales-data",
"Train.csv"
)

# Load test data

print("Loading test data from Kaggle...")
test = kagglehub.load_dataset(
KaggleDatasetAdapter.PANDAS,
"brijbhushannanda1979/bigmart-sales-data",
"Test.csv"
)

print("Data loaded successfully from Kaggle!")

except Exception as e:
print(f"Error loading data from Kaggle: {e}")
print("Falling back to URL loading...")

try:
# Fallback to GitHub URLs
train_url = "https://raw.githubusercontent.com/suvikramsain/Bigmart-Sales/master/Train.csv"
test_url = "https://raw.githubusercontent.com/suvikramsain/Bigmart-Sales/master/Test.csv"

train = pd.read_csv(train_url)
test = pd.read_csv(test_url)
print("Data loaded successfully from GitHub URLs")
except:
print("Error loading data from URLs. Using local files if available.")
try:
train = pd.read_csv('Train.csv')
test = pd.read_csv('Test.csv')
print("Data loaded successfully from local files")
except:

https://colab.research.google.com/drive/1nQacKbOQQd-CivRD_lxjOwG0_1Eg0wf0#scrollTo=dx1pZ_GLtDdh&printMode=true 1/5
4/22/25, 1:36 AM Untitled4.ipynb - Colab
print("Error: Unable to load data. Please check data availability.")
import sys
sys.exit(1)

Loading train data from Kaggle...

Downloading from https://www.kaggle.com/api/v1/datasets/download/brijbhushannanda1979/bigmart-sales-data?dataset_version_number=1&fi
100%|██████████| 849k/849k [00:00<00:00, 70.8MB/s]Loading test data from Kaggle...

Data loaded successfully from Kaggle!

 

# Take a look at the data

print("Train data shape:", train.shape)
print("Test data shape:", test.shape)
print("\nTrain data columns:", train.columns.tolist())
print("\nFirst few rows of train data:")
print(train.head())

Train data shape: (8523, 12)

Test data shape: (5681, 11)

Train data columns: ['Item_Identifier', 'Item_Weight', 'Item_Fat_Content', 'Item_Visibility', 'Item_Type', 'Item_MRP', 'Outlet_Ident

First few rows of train data:

Item_Identifier Item_Weight Item_Fat_Content Item_Visibility \
0 FDA15 9.30 Low Fat 0.016047
1 DRC01 5.92 Regular 0.019278
2 FDN15 17.50 Low Fat 0.016760
3 FDX07 19.20 Regular 0.000000
4 NCD19 8.93 Low Fat 0.000000

Item_Type Item_MRP Outlet_Identifier \

0 Dairy 249.8092 OUT049
1 Soft Drinks 48.2692 OUT018
2 Meat 141.6180 OUT049
3 Fruits and Vegetables 182.0950 OUT010
4 Household 53.8614 OUT013

Outlet_Establishment_Year Outlet_Size Outlet_Location_Type \

0 1999 Medium Tier 1
1 2009 Medium Tier 3
2 1999 Medium Tier 1
3 1998 NaN Tier 3
4 1987 High Tier 3

Outlet_Type Item_Outlet_Sales
0 Supermarket Type1 3735.1380
1 Supermarket Type2 443.4228
2 Supermarket Type1 2097.2700
3 Grocery Store 732.3800
4 Supermarket Type1 994.7052

 

# Check for missing values

print("\nMissing values in train data:")
print(train.isnull().sum())
print("\nMissing values in test data:")
print(test.isnull().sum())

Missing values in train data:

Item_Identifier 0
Item_Weight 1463
Item_Fat_Content 0
Item_Visibility 0
Item_Type 0
Item_MRP 0
Outlet_Identifier 0
Outlet_Establishment_Year 0
Outlet_Size 2410
Outlet_Location_Type 0
Outlet_Type 0
Item_Outlet_Sales 0
dtype: int64

Missing values in test data:

Item_Identifier 0
Item_Weight 976
Item_Fat_Content 0
Item_Visibility 0
Item_Type 0

https://colab.research.google.com/drive/1nQacKbOQQd-CivRD_lxjOwG0_1Eg0wf0#scrollTo=dx1pZ_GLtDdh&printMode=true 2/5
4/22/25, 1:36 AM Untitled4.ipynb - Colab
Item_MRP 0
Outlet_Identifier 0
Outlet_Establishment_Year 0
Outlet_Size 1606
Outlet_Location_Type 0
Outlet_Type 0
dtype: int64

# Combine train and test datasets for preprocessing

test['Item_Outlet_Sales'] = np.nan
combined = pd.concat([train, test], ignore_index=True)
print("\nCombined data shape:", combined.shape)

# Data preprocessing
# Fill missing values
combined['Item_Weight'].fillna(combined['Item_Weight'].mean(), inplace=True)
combined['Outlet_Size'].fillna('Unknown', inplace=True)

# Fix inconsistent categories

combined['Item_Fat_Content'] = combined['Item_Fat_Content'].replace({
'LF': 'Low Fat',
'low fat': 'Low Fat',
'reg': 'Regular'
})

Combined data shape: (14204, 12)

# Feature Engineering
# Extract year feature from establishment year
current_year = 2025
combined['Outlet_Years'] = current_year - combined['Outlet_Establishment_Year']
combined.drop('Outlet_Establishment_Year', axis=1, inplace=True)

# Item visibility should not be 0, replace with mean

zero_visibility_mask = combined['Item_Visibility'] == 0
combined.loc[zero_visibility_mask, 'Item_Visibility'] = combined['Item_Visibility'].mean()

# Normalize Item_Visibility
combined['Item_Visibility'] = combined['Item_Visibility'] / combined['Item_Visibility'].max()

# Encode categorical variables

le = LabelEncoder()
for column in ['Item_Fat_Content', 'Item_Type', 'Outlet_Size', 'Outlet_Location_Type', 'Outlet_Type']:
combined[column] = le.fit_transform(combined[column])

# Create dummy variables for categorical features

categorical_columns = ['Item_Identifier', 'Outlet_Identifier']
combined = pd.get_dummies(combined, columns=categorical_columns, drop_first=True)

# Split back into train and test

train_processed = combined[~combined['Item_Outlet_Sales'].isnull()]
test_processed = combined[combined['Item_Outlet_Sales'].isnull()]

# Prepare data for modeling

X = train_processed.drop('Item_Outlet_Sales', axis=1)
y = train_processed['Item_Outlet_Sales']
X_test = test_processed.drop('Item_Outlet_Sales', axis=1)

# Split into training and validation sets

X_train, X_val, y_train, y_val = train_test_split(X, y, test_size=0.2, random_state=42)

# Scale the numerical features

scaler = StandardScaler()
numerical_columns = ['Item_Weight', 'Item_Visibility', 'Item_MRP', 'Outlet_Years']
X_train[numerical_columns] = scaler.fit_transform(X_train[numerical_columns])
X_val[numerical_columns] = scaler.transform(X_val[numerical_columns])
X_test[numerical_columns] = scaler.transform(X_test[numerical_columns])

# Train a Random Forest Regressor

print("\nTraining Random Forest Regressor...")
rf = RandomForestRegressor(n_estimators=100, max_depth=10, random_state=42)
rf.fit(X_train, y_train)

https://colab.research.google.com/drive/1nQacKbOQQd-CivRD_lxjOwG0_1Eg0wf0#scrollTo=dx1pZ_GLtDdh&printMode=true 3/5
4/22/25, 1:36 AM Untitled4.ipynb - Colab

Training Random Forest Regressor...

▾ RandomForestRegressor i ?

RandomForestRegressor(max_depth=10, random_state=42)

# Evaluate the model

y_pred_val = rf.predict(X_val)
rmse_val = np.sqrt(mean_squared_error(y_val, y_pred_val))
print(f"Validation RMSE: {rmse_val:.4f}")

# Feature importance
feature_importances = pd.DataFrame({
'Feature': X_train.columns,
'Importance': rf.feature_importances_
}).sort_values('Importance', ascending=False)

print("\nTop 10 most important features:")

print(feature_importances.head(10))

# Make predictions on test data

y_pred_test = rf.predict(X_test)

# Create submission file

submission = pd.DataFrame({
'Item_Identifier': test['Item_Identifier'],
'Outlet_Identifier': test['Outlet_Identifier'],
'Item_Outlet_Sales': y_pred_test
})

# Save the submission file

submission.to_csv('bigmart_sales_prediction.csv', index=False)
print("\nSubmission file created successfully!")

# Visualize actual vs predicted values on validation set

plt.figure(figsize=(10, 6))
plt.scatter(y_val, y_pred_val, alpha=0.5)
plt.plot([y_val.min(), y_val.max()], [y_val.min(), y_val.max()], 'r--')
plt.xlabel('Actual Sales')
plt.ylabel('Predicted Sales')
plt.title('Actual vs Predicted Sales')
plt.tight_layout()
plt.savefig('actual_vs_predicted.png')

# Visualize feature importance

plt.figure(figsize=(12, 8))
sns.barplot(x='Importance', y='Feature', data=feature_importances.head(15))
plt.title('Top 15 Feature Importance')
plt.tight_layout()
plt.savefig('feature_importance.png')

print("\nAnalysis completed!")

https://colab.research.google.com/drive/1nQacKbOQQd-CivRD_lxjOwG0_1Eg0wf0#scrollTo=dx1pZ_GLtDdh&printMode=true 4/5
4/22/25, 1:36 AM Untitled4.ipynb - Colab

Validation RMSE: 1027.9054

Top 10 most important features:

Feature Importance
4 Item_MRP 0.498320
7 Outlet_Type 0.314966
8 Outlet_Years 0.042362
1571 Outlet_Identifier_OUT027 0.033873
2 Item_Visibility 0.015169
0 Item_Weight 0.005936
3 Item_Type 0.005011
1318 Item_Identifier_NCE42 0.003391
796 Item_Identifier_FDQ19 0.002214
1218 Item_Identifier_FDY55 0.002083

Submission file created successfully!

Analysis completed!

https://colab.research.google.com/drive/1nQacKbOQQd-CivRD_lxjOwG0_1Eg0wf0#scrollTo=dx1pZ_GLtDdh&printMode=true 5/5