# This Python 3 environment comes with many helpful analytics libraries installed

# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python

# For example, here's several helpful packages to load

import numpy as np # linear algebra

import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)

# Input data files are available in the read-only "../input/" directory

# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory

import os
for dirname, _, filenames in os.walk('/kaggle/input'):
for filename in filenames:
print(os.path.join(dirname, filename))

# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create
# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session

/kaggle/input/student-bag-price-prediction-dataset/Noisy_Student_Bag_Price_Prediction_Dataset.csv
/kaggle/input/playground-series-s5e2/sample_submission.csv
/kaggle/input/playground-series-s5e2/train.csv
/kaggle/input/playground-series-s5e2/test.csv
/kaggle/input/playground-series-s5e2/training_extra.csv

import pandas as pd, numpy as np

train = pd.read_csv("/kaggle/input/playground-series-s5e2/train.csv")
print("Train shape",train.shape)
train_extra = pd.read_csv("/kaggle/input/playground-series-s5e2/training_extra.csv")
print("Extra Train shape",train_extra.shape)
train = pd.concat([train,train_extra],axis=0,ignore_index=True)
print("Combined Train shape",train.shape)

Train shape (300000, 11)

Extra Train shape (3694318, 11)
Combined Train shape (3994318, 11)

train.head(10)

Laptop Weight
id Brand Material Size Compartments Waterproof Style Color Price
Compartment Capacity (kg)

0 0 Jansport Leather Medium 7.0 Yes No Tote Black 11.611723 112.15875

1 1 Jansport Canvas Small 10.0 Yes Yes Messenger Green 27.078537 68.88056

Under
2 2 Leather Small 2.0 Yes No Messenger Red 16.643760 39.17320
Armour

3 3 Nike Nylon Small 8.0 Yes No Messenger Green 12.937220 80.60793

4 4 Adidas Canvas Medium 1.0 Yes Yes Messenger Green 17.749338 86.02312

5 5 Nike Canvas Medium 10.0 No Yes NaN Black 7.241812 20.01553

6 6 Nike NaN Large 3.0 No No Backpack Green 6.828123 84.80500

7 7 Puma Canvas Small 1.0 Yes Yes Backpack Blue 21.488864 27.15815

Under
8 8 Polyester Medium 8.0 Yes No Tote Gray 10.207780 25.98652
Armour

Under
9 9 Nylon Medium 2.0 Yes Yes Messenger Pink 15.895100 38.48741
Armour

# Assuming your DataFrame is named df

unique_values = train['Weight Capacity (kg)'].unique()
print("Unique Weight Capacity values:", unique_values)

Unique Weight Capacity values: [11.61172281 27.07853658 16.64375995 ... 12.79080004 22.95972519
16.64173875]

# Assuming your DataFrame is named df

unique_values = train['Weight Capacity (kg)'].value_counts()
print("Unique Weight Capacity values:", unique_values)
Unique Weight Capacity values: Weight Capacity (kg)
5.000000 58087
30.000000 2588
11.898250 1571
14.908437 1559
22.898382 1417
...
20.923873 1
28.897699 1
20.485038 1
14.355831 1
21.643489 1
Name: count, Length: 1920345, dtype: int64

import matplotlib.pyplot as plt

import seaborn as sns

plt.figure(figsize=(10, 6))
sns.histplot(train['Weight Capacity (kg)'], bins=30, kde=False)
plt.title("Distribution of Weight Capacity")
plt.xlabel("Weight Capacity")
plt.ylabel("Count")
plt.show()

/usr/local/lib/python3.10/dist-packages/seaborn/_oldcore.py:1119: FutureWarning: use_inf_as_na option is depreca

ted and will be removed in a future version. Convert inf values to NaN before operating instead.
with pd.option_context('mode.use_inf_as_na', True):

train_mean = train.Price.mean()
train['pred'] = train_mean
s = np.sqrt(np.mean( (train.Price-train.pred)**2.0 ) )
print(f"Validation RMSE using Train Mean = {s}")

Validation RMSE using Train Mean = 38.93867923358143

train.head()

Weight
Laptop
id Brand Material Size Compartments Waterproof Style Color Capacity Price pred
Compartment
(kg)

0 0 Jansport Leather Medium 7.0 Yes No Tote Black 11.611723 112.15875 81.362175

1 1 Jansport Canvas Small 10.0 Yes Yes Messenger Green 27.078537 68.88056 81.362175

Under
2 2 Leather Small 2.0 Yes No Messenger Red 16.643760 39.17320 81.362175
Armour

3 3 Nike Nylon Small 8.0 Yes No Messenger Green 12.937220 80.60793 81.362175

4 4 Adidas Canvas Medium 1.0 Yes Yes Messenger Green 17.749338 86.02312 81.362175

from cuml.preprocessing import TargetEncoder

 TE = TargetEncoder(n_folds=30, smooth=20, split_method='random', stat='mean')
train['pred'] = TE.fit_transform(train['Weight Capacity (kg)'],train.Price)
s = np.sqrt(np.mean( (train.Price-train.pred)**2.0 ) )
print(f"Validation RSME using Target Encode Weight Capacity = {s}")

Validation RSME using Target Encode Weight Capacity = 38.71037277159231

from cuml.preprocessing import TargetEncoder

import numpy as np
import pandas as pd

# Variation 1: More smoothing

TE1 = TargetEncoder(n_folds=30, smooth=40, split_method='random', stat='mean')
train['pred_TE1'] = TE1.fit_transform(train['Weight Capacity (kg)'], train.Price)
rmse_TE1 = np.sqrt(np.mean((train.Price - train['pred_TE1'])**2.0))
print(f"Variation 1 (smooth=40) RMSE: {rmse_TE1}")

# Variation 2: Fewer folds and less smoothing

TE2 = TargetEncoder(n_folds=10, smooth=10, split_method='random', stat='mean')
train['pred_TE2'] = TE2.fit_transform(train['Weight Capacity (kg)'], train.Price)
rmse_TE2 = np.sqrt(np.mean((train.Price - train['pred_TE2'])**2.0))
print(f"Variation 2 (n_folds=10, smooth=10) RMSE: {rmse_TE2}")

Variation 1 (smooth=40) RMSE: 38.719630133491286

Variation 2 (n_folds=10, smooth=10) RMSE: 38.729537426961386

from sklearn.model_selection import KFold

import numpy as np

# Create a KFold instance

kf = KFold(n_splits=30, shuffle=True, random_state=42)

# Initialize an array to hold fold IDs for each sample

fold_ids = np.empty(len(train), dtype=int)

# Assign fold IDs

for fold, (_, val_idx) in enumerate(kf.split(train)):
fold_ids[val_idx] = fold

# Now use your custom fold_ids with the target encoder

TE3 = TargetEncoder(n_folds=30, smooth=20, split_method='customize', stat='mean')
train['pred_TE3'] = TE3.fit_transform(train['Weight Capacity (kg)'], train.Price, fold_ids=fold_ids)

rmse_TE3 = np.sqrt(np.mean((train.Price - train['pred_TE3'])**2.0))

print(f"Variation 2 (n_folds=30, smooth=20) RMSE: {rmse_TE3}")

Variation 2 (n_folds=30, smooth=20) RMSE: 38.71077699918216

# Now you can use these features in a model, e.g., a simple linear regression:
from sklearn.linear_model import LinearRegression
features = ['pred_TE1', 'pred_TE2', 'pred_TE3']
lr_model = LinearRegression()
lr_model.fit(train[features], train.Price)

# Evaluate RMSE on the training set (or better, via cross-validation)

pred_lr = lr_model.predict(train[features])
rmse_lr = np.sqrt(np.mean((train.Price - pred_lr)**2.0))
print(f"Linear Regression using multiple target encoding features RMSE: {rmse_lr}")

Linear Regression using multiple target encoding features RMSE: 38.70900752709825

test = pd.read_csv("/kaggle/input/playground-series-s5e2/test.csv")

# Step 1: Apply Target Encoding to Test Set

test['pred_TE1'] = TE1.transform(test['Weight Capacity (kg)'])
test['pred_TE2'] = TE2.transform(test['Weight Capacity (kg)'])
test['pred_TE3'] = TE3.transform(test['Weight Capacity (kg)'])

# Step 2: Predict on Test Set using Linear Regression Model

test_features = ['pred_TE1', 'pred_TE2', 'pred_TE3']
test['Price'] = lr_model.predict(test[test_features])

# Step 3: Prepare Submission File

sub = pd.read_csv("/kaggle/input/playground-series-s5e2/sample_submission.csv") # Load sample submission
sub['Price'] = test['Price']
sub.to_csv("submission_lr_target_encoding.csv", index=False) # Save as CSV

# Step 4: Verify Submission File

print(sub.head()) # Check the first few rows to ensure everything looks correct
 id Price
0 300000 82.787024
1 300001 81.064362
2 300002 90.514901
3 300003 78.539448
4 300004 81.464757

# sub = pd.read_csv("/kaggle/input/playground-series-s5e2/sample_submission.csv")
# print('Submission shape', sub.shape)
# test = pd.read_csv("/kaggle/input/playground-series-s5e2/test.csv")
# sub['Price'] = TE.transform(test['Weight Capacity (kg)'])
# sub.to_csv("submission_TE_weight_capacity.csv",index=False)
# sub.head()