souce_code.py

# -*- coding: utf-8 -*-
"""Mini_Project.ipynb

Automatically generated by Colab.

Original file is located at
    https://colab.research.google.com/drive/1RJPzowjB2WGLAwpwDOnAmyBnNWe5J5id
"""

import pandas as pd

df = pd.read_csv('/content/ps5.csv')

# Display the first few rows of the dataset
print(df.head())

# Get summary statistics
print(df.describe())

# Get information about the dataset
print(df.info())

# Check for missing values
print(df.isnull().sum())

# Drop or fill missing values
df = df.dropna()  # or use df.fillna(method='ffill') for forward filling

# Check for duplicates
df = df.drop_duplicates()

# Inspect the releaseDate column
print(df['releaseDate'].head())
print(df['releaseDate'].dtype)

# Convert releaseDate to datetime
df['releaseDate'] = pd.to_datetime(df['releaseDate'], errors='coerce')

# Check for any conversion issues
print(df['releaseDate'].isnull().sum())  # This will show how many entries could not be converted

# Extract year and month from releaseDate
df['releaseYear'] = df['releaseDate'].dt.year
df['releaseMonth'] = df['releaseDate'].dt.month

# Verify the new columns
print(df[['releaseDate', 'releaseYear', 'releaseMonth']].head())

import seaborn as sns
import matplotlib.pyplot as plt

# Plot user ratings over the years
plt.figure(figsize=(10, 6))
sns.lineplot(x='releaseYear', y='starRating/averageRating', data=df)
plt.title('User Ratings Over the Years')
plt.xlabel('Release Year')
plt.ylabel('Average Rating')
plt.show()

# Distribution of user ratings by release month
plt.figure(figsize=(10, 6))
sns.boxplot(x='releaseMonth', y='starRating/averageRating', data=df)
plt.title('User Ratings by Release Month')
plt.xlabel('Release Month')
plt.ylabel('Average Rating')
plt.show()

# Top publishers by average rating
top_publishers = df.groupby('publisherName')['starRating/averageRating'].mean().sort_values(ascending=False).head(10)
print(top_publishers)

# Plot average ratings for top publishers
plt.figure(figsize=(12, 6))
sns.barplot(x=top_publishers.index, y=top_publishers.values)
plt.title('Top Publishers by Average Rating')
plt.xlabel('Publisher')
plt.ylabel('Average Rating')
plt.xticks(rotation=45)
plt.show()

import pandas as pd

# Example DataFrame
data = {
    'url': ['http://example.com', 'http://example2.com'],
    'id': ['1', '2'],
    'publisherName': ['Publisher1', 'Unknown'],
    'releaseDate': ['2022-01-01', '2022-02-01'],
    'name': ['Item1', 'Item2'],
    'isAgeRestricted': [True, False],
    'activeCtaId': ['cta1', 'cta2'],
    'starRating/averageRating': [4.5, 3.8],
    'starRating/totalRatingsCount': [100, 50]
}

df = pd.DataFrame(data)

# Average rating by age restriction
age_restrictions = df.groupby('isAgeRestricted')['starRating/averageRating'].mean()
print(age_restrictions)

# Optionally, if you want to visualize the result:
import matplotlib.pyplot as plt

age_restrictions.plot(kind='bar', color=['blue', 'orange'])
plt.xlabel('Is Age Restricted')
plt.ylabel('Average Star Rating')
plt.title('Average Star Rating by Age Restriction')
plt.xticks(ticks=[0, 1], labels=['False', 'True'], rotation=0)
plt.show()

print(df.columns)

# Check for string values in features
print(df.select_dtypes(include=['object']).columns)

# Check for string values in target variable
print(df['starRating/averageRating'].dtype)

import joblib
import pandas as pd

# Load the pre-trained model
model = joblib.load('ps5_game_rating_predictor.pkl')

# Example new data with sample values
new_data = {
    'url': ['https://store.playstation.com/en-us/product/UP0101-PPSA19225_00-0159266583099383'],
    'id': ['1'],
    'publisherName': ['Konami Digital Entertainment, Inc.'],  # Sample publisher
    'releaseDate': ['2024-07-20'],  # Sample release date
    'name': ['Sample Game'],  # Sample game name
    'isAgeRestricted': [False],  # Sample age restriction
    'activeCtaId': ['cta1'],  # Sample activeCtaId
    'starRating/totalRatingsCount': [50]  # Sample total ratings count
}

# Convert new data to DataFrame
new_df = pd.DataFrame(new_data)

# Drop non-numeric columns that were not used in training
non_numeric_cols = ['url', 'id', 'publisherName', 'name', 'releaseDate', 'activeCtaId']
new_df.drop(columns=non_numeric_cols, inplace=True)

# Ensure the correct dtype for isAgeRestricted
new_df['isAgeRestricted'] = new_df['isAgeRestricted'].astype(int)  # Convert boolean to int

# Manually specify the features used during training
# Ensure this list matches exactly the features used during training
training_features = [
    'isAgeRestricted'
]

# Ensure new_df has the same columns as used during training
new_df = new_df[training_features]

# Make prediction
prediction = model.predict(new_df)

# Print the prediction
print(f'Predicted Average Rating: {prediction[0]}')

#correctThisOne