import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt

df = pd.read_csv('vgsales.csv')

df.head()

df.head()

df.tail()

df.sample()

df.shape

data_types = df.dtypes
print(data_types)

print("\nMissing values in the dataset:")
print(df.isnull().sum())

# Remove rows where 'Name' or 'Genre' is missing
df.dropna(subset=['Name', 'Genre'], inplace=True)

# Replace missing 'Publisher' values with 'Unknown'
df['Publisher'].fillna('Unknown', inplace=True)

# Display the DataFrame to confirm changes
df.head()

print("\nStatistical details of the dataset:")
print(df.describe())

# Exclude 'Year_of_Release' from the statistical summary
statistical_details = df.drop(columns='Year_of_Release').describe()

# Print the statistical details of the dataset excluding 'Year_of_Release'
print("\nStatistical details of the dataset (excluding 'Year_of_Release'):")
print(statistical_details)

platform_counts = df['Platform'].value_counts()
print("\nNumber of games per platform:")
print(platform_counts)

genre_counts = df['Genre'].value_counts()
print("\nNumber of games per genre:")
print(genre_counts)

# Define the platform to console brand mapping
platform_brand_mapping = {
    'Wii': 'Nintendo', 'NES': 'Nintendo', 'GB': 'Nintendo', 'DS': 'Nintendo', 'SNES': 'Nintendo',
    '3DS': 'Nintendo', 'N64': 'Nintendo', 'GBA': 'Nintendo', 'GC': 'Nintendo', 'WiiU': 'Nintendo', 'Switch': 'Nintendo',
    'PS': 'Sony', 'PS2': 'Sony', 'PS3': 'Sony', 'PS4': 'Sony', 'PSP': 'Sony', 'PSV': 'Sony',
    'X360': 'Microsoft', 'XB': 'Microsoft', 'XOne': 'Microsoft',
    'GEN': 'Sega', 'DC': 'Sega', 'SAT': 'Sega', 'SCD': 'Sega',
    'WS': 'Bandai', 'NG': 'SNK', 'TG16': 'NEC', '3DO': '3DO Company', 'GG': 'Sega', 'PCFX': 'NEC',
    'PC': 'PC'  # PC gaming
}

# Create or update the Console_Brand column based on the mapping
df['Console_Brand'] = df['Platform'].map(platform_brand_mapping)


unmapped_platforms = df[df['Console_Brand'].isnull()]['Platform'].unique()
if len(unmapped_platforms) > 0:
    print("Unmapped Platforms:", unmapped_platforms)


print(df.isnull().sum())

# Use the mapping to fill missing Console_Brand values based on Platform
df['Console_Brand'] = df['Console_Brand'].fillna(df['Platform'].map(platform_brand_mapping))

print(df.isnull().sum())

# Identify platforms with missing Console_Brand values
unmapped_platforms = df[df['Console_Brand'].isnull()]['Platform'].unique()
print("Unmapped Platforms:", unmapped_platforms)

print(df.isnull().sum())

# Calculate the percentage of global sales for each region and create new columns
df['EU_Sales_Percent'] = (df['EU_Sales'] / df['Global_Sales']) * 100
df['NA_Sales_Percent'] = (df['NA_Sales'] / df['Global_Sales']) * 100
df['JP_Sales_Percent'] = (df['JP_Sales'] / df['Global_Sales']) * 100

# Display the updated DataFrame with new percentage columns
print(df[['Name', 'EU_Sales_Percent', 'NA_Sales_Percent', 'JP_Sales_Percent']].head())

# Calculate genre popularity for Japan
jp_genre_popularity = df.groupby('Genre')['JP_Sales'].sum().sort_values(ascending=False).reset_index()
jp_genre_popularity.rename(columns={'JP_Sales': 'JP_Genre_Popularity'}, inplace=True)

# Calculate genre popularity for Europe
eu_genre_popularity = df.groupby('Genre')['EU_Sales'].sum().sort_values(ascending=False).reset_index()
eu_genre_popularity.rename(columns={'EU_Sales': 'EU_Genre_Popularity'}, inplace=True)

jp_genre_popularity, eu_genre_popularity

df.head()

df.head(150)

import pandas as pd

# Calculate total global sales per publisher
publisher_sales = df.groupby('Publisher')['Global_Sales'].sum()

# Count the number of games published by each publisher
publisher_game_count = df.groupby('Publisher').size()

# Calculate the sales performance score as total sales divided by the number of games
publisher_performance_score = publisher_sales / publisher_game_count

# Create a DataFrame for the scores
publisher_performance = pd.DataFrame({
    'Total_Sales': publisher_sales,
    'Game_Count': publisher_game_count,
    'Sales_Performance_Score': publisher_performance_score
}).reset_index()

# Display the publishers with the highest performance scores
print(publisher_performance.sort_values(by='Sales_Performance_Score', ascending=False).head(10))

# Sum of global sales by genre
sales_by_genre = df.groupby('Genre')['Global_Sales'].sum().sort_values(ascending=False)

plt.figure(figsize=(12, 8))
sns.barplot(x=sales_by_genre.values, y=sales_by_genre.index)
plt.title('Global Sales by Genre')
plt.xlabel('Global Sales (in millions)')
plt.ylabel('Genre')
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt


# Sum of global sales by genre, sorted in descending order
sales_by_genre = df.groupby('Genre')['Global_Sales'].sum().sort_values(ascending=False)

# Setting the figure size for better readability
plt.figure(figsize=(12, 8))

# Creating a barplot; sns.barplot automatically sorts the bars in descending order
sns.barplot(x=sales_by_genre.values, y=sales_by_genre.index, palette='viridis')  # Using a visually appealing color palette

plt.title('Global Sales by Genre', fontsize=20)
plt.xlabel('Global Sales (in millions)', fontsize=14)
plt.ylabel('Genre', fontsize=14)

plt.grid(axis='x', linestyle='--', alpha=0.6)

# Displaying the plot
plt.show()

# Sort the DataFrame by 'Global_Sales' select the top 5
top_5_games = df.sort_values(by='Global_Sales', ascending=False).head(5)

# Display the top 5 games
print(top_5_games[['Name', 'Global_Sales']])

plt.figure(figsize=(12, 6))
sns.barplot(x='Global_Sales', y='Name', data=top_5_games, palette='viridis')
plt.title('Top 5 Games by Global Sales')
plt.xlabel('Global Sales (in millions)')
plt.ylabel('Game Name')
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt


plt.figure(figsize=(12, 6))

# Creating a barplot with a visually appealing color palette
sns.barplot(x='Global_Sales', y='Name', data=top_5_games, palette='viridis')

# Adding a descriptive title and axis labels with specified font sizes
plt.title('Top 5 Games by Global Sales', fontsize=18)
plt.xlabel('Global Sales (in millions)', fontsize=14)
plt.ylabel('Game Name', fontsize=14)

# Adding horizontal grid lines for better readability of sales values
plt.grid(axis='x', linestyle='--', alpha=0.7)

# Optional: Annotating the bars with the exact sales values for precise information
for index, value in enumerate(top_5_games['Global_Sales']):
    plt.text(value, index, f' {value:.2f}', va='center', fontsize=12, color='black')

# Displaying the plot
plt.show()

# Convert 'Year_of_Release' to string for concatenation
top_5_games['Year_of_Release'] = top_5_games['Year_of_Release'].astype(str)

# Create a new column 'Name_Year' that combines 'Name' and 'Year_of_Release'
top_5_games['Name_Year'] = top_5_games['Name'] + ' (' + top_5_games['Year_of_Release'] + ')'

plt.figure(figsize=(12, 6))
sns.barplot(x='Global_Sales', y='Name_Year', data=top_5_games, palette='viridis')
plt.title('Top 5 Games by Global Sales')
plt.xlabel('Global Sales (in millions)')
plt.ylabel('Game Name (Year of Release)')
plt.show()

#Enchanced

# Setting the figure size for a clear and spacious presentation
plt.figure(figsize=(12, 6))

# Creating a bar plot with a visually appealing color palette
sns.barplot(x='Global_Sales', y='Name_Year', data=top_5_games, palette='viridis')

# Adding a descriptive title and specifying font sizes for axis labels
plt.title('Top 5 Games by Global Sales', fontsize=18)
plt.xlabel('Global Sales (in millions)', fontsize=14)
plt.ylabel('Game Name (Year of Release)', fontsize=14)

# Adding horizontal grid lines to enhance readability
plt.grid(axis='x', linestyle='--', alpha=0.7)

# Annotating each bar with the exact global sales figure for precise information
for index, value in enumerate(top_5_games['Global_Sales']):
    plt.text(value, index, f' {value:.2f}M', va='center', fontsize=12, color='black')

# Display the plot
plt.show()

print(df.columns)

# Filter the DataFrame for the years 2010-2014
df_filtered = df[(df['Year_of_Release'] >= 2010) & (df['Year_of_Release'] <= 2014)]

# Calculate the correlation coefficient between 'NA_Sales' and 'JP_Sales'
correlation = df_filtered['NA_Sales'].corr(df_filtered['JP_Sales'])

print(f"The correlation between NA sales and JP sales for the years 2010-2014 is: {correlation}")

import matplotlib.pyplot as plt
import seaborn as sns

palette = sns.cubehelix_palette(start=.5, rot=-.75, as_cmap=True)

# Create a scatter plot with a regression line
plt.figure(figsize=(10, 6))

sns.regplot(x='NA_Sales', y='JP_Sales', data=df_filtered,
            scatter_kws={'alpha':0.6, 'cmap': palette}, line_kws={'color':'#2ca02c'},
            scatter=True, fit_reg=True)

plt.title('Relationship Between NA Sales and JP Sales (2010-2014)')
plt.xlabel('NA Sales (in millions)')
plt.ylabel('JP Sales (in millions)')

norm = plt.Normalize(df_filtered['NA_Sales'].min(), df_filtered['NA_Sales'].max())
sm = plt.cm.ScalarMappable(cmap=palette, norm=norm)
sm.set_array([])
plt.colorbar(sm, label='NA Sales Density')

plt.show()

plt.figure(figsize=(10, 6))
# Assuming 'Genre' is a relevant variable
sns.scatterplot(x='NA_Sales', y='JP_Sales', data=df_filtered, hue='Genre', alpha=0.5, palette='Set1')
plt.title('Colored Relationship Between NA Sales and JP Sales by Genre (2010-2014)')
plt.xlabel('NA Sales (in millions)')
plt.ylabel('JP Sales (in millions)')
plt.legend(title='Genre', bbox_to_anchor=(1.05, 1), loc='upper left')
plt.show()

top_genres_df['Log_Global_Sales'] = np.log1p(top_genres_df['Global_Sales'])

plt.figure(figsize=(10, 6))
sns.violinplot(x='Genre', y='Log_Global_Sales', data=top_genres_df, palette='Set2')

plt.title('Log-Transformed Distribution of Global Sales for Top 4 Game Genres')
plt.xlabel('Genre')
plt.ylabel('Log of Global Sales (in millions)')

plt.show()

import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np


# Applying log transformation to Global Sales using np.log1p for better numerical stability
top_genres_df['Log_Global_Sales'] = np.log1p(top_genres_df['Global_Sales'])

# Setting the figure size
plt.figure(figsize=(10, 6))

# Creating a violin plot to show the distribution of log-transformed global sales across different genres
sns.violinplot(x='Genre', y='Log_Global_Sales', data=top_genres_df, palette='Set2', inner='quartile')

# Customizing the plot with a title, and labels for x and y axes
plt.title('Log-Transformed Distribution of Global Sales for Top 4 Game Genres', fontsize=16)
plt.xlabel('Genre', fontsize=14)
plt.ylabel('Log of Global Sales (in millions)', fontsize=14)

# Improving the layout to prevent label cutoff and displaying the plot
plt.tight_layout()
plt.show()

import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np


top_genres_df['Log_Global_Sales'] = np.log1p(top_genres_df['Global_Sales'])

# Setting the figure size
plt.figure(figsize=(10, 6))

# Creating a violin plot to show the distribution of log-transformed global sales across different genres
sns.violinplot(x='Genre', y='Log_Global_Sales', data=top_genres_df, palette='Set2', inner='quartile')

# Customizing the plot with a title, and labels for x and y axes
plt.title('Log-Transformed Distribution of Global Sales for Top 4 Game Genres', fontsize=16)
plt.xlabel('Genre', fontsize=14)
plt.ylabel('Log of Global Sales (in millions)', fontsize=14)

# Improving the layout to prevent label cutoff and displaying the plot
plt.tight_layout()
plt.show()

plt.figure(figsize=(12, 6))
sns.stripplot(x='Genre', y='Log_Global_Sales', data=top_genres_df, palette='Set2', jitter=True, alpha=0.5)
plt.title('Strip Plot of Log-Transformed Global Sales for Top 4 Game Genres')
plt.xlabel('Genre')
plt.ylabel('Log of Global Sales (in millions)')
plt.show()

plt.figure(figsize=(12, 6))
sns.pointplot(x='Genre', y='Log_Global_Sales', data=top_genres_df, capsize=.2, palette='Set2')
plt.title('Point Plot of Log-Transformed Global Sales for Top 4 Game Genres')
plt.xlabel('Genre')
plt.ylabel('Log of Global Sales (in millions)')
plt.show()

g = sns.FacetGrid(top_genres_df, col='Genre', col_wrap=2, height=4, aspect=1.5)
g.map(sns.histplot, 'Log_Global_Sales', kde=True, bins=15, color='skyblue')
g.add_legend()
g.set_titles('{col_name} Genre')
g.set_axis_labels('Log of Global Sales (in millions)', 'Count')
plt.show()

# Group the dataset into older and newer games
older_games = df[df['Year_of_Release'] <= 2005]
newer_games = df[df['Year_of_Release'] > 2005]

# Calculate the mean EU sales for each group
mean_eu_sales_older = older_games['EU_Sales'].mean()
mean_eu_sales_newer = newer_games['EU_Sales'].mean()

print(f"Mean EU Sales for Older Games (2005 and earlier): {mean_eu_sales_older:.2f}")
print(f"Mean EU Sales for Newer Games (after 2005): {mean_eu_sales_newer:.2f}")

# Compare the means
if mean_eu_sales_older > mean_eu_sales_newer:
    print("Older games (2005 and earlier) have higher mean EU sales than newer games.")
elif mean_eu_sales_older < mean_eu_sales_newer:
    print("Newer games (after 2005) have higher mean EU sales than older games.")
else:
    print("Mean EU sales are the same for older and newer games.")

import matplotlib.pyplot as plt
import seaborn as sns

# Data for plotting
categories = ['Games (≤2005)', 'Games (>2005)']
mean_sales = [mean_eu_sales_older, mean_eu_sales_newer]

plt.figure(figsize=(8, 6))

# Create a bar plot
sns.barplot(x=categories, y=mean_sales, palette='coolwarm')

# Add titles and labels
plt.title('Comparison of Mean EU Sales')
plt.ylabel('Mean EU Sales (in millions)')
plt.xlabel('Game Category')

# Display the values on the bars
for i, value in enumerate(mean_sales):
    plt.text(i, value + 0.01, f"{value:.2f}", ha='center', va='bottom')

plt.show()

plt.figure(figsize=(8, 6))
sns.barplot(x=categories, y=mean_sales, palette='coolwarm')
plt.ylim(0.13, 0.16)  # Adjust the limits based on your data to zoom in
plt.title('Comparison of Mean EU Sales for Older vs. Newer Games')
plt.ylabel('Mean EU Sales (in millions)')
plt.xlabel('Game Category')
plt.show()

plt.figure(figsize=(8, 6))
sns.barplot(x=categories, y=mean_sales, palette='coolwarm')

plt.ylim(0.14, 0.155)

plt.title('Comparison of Mean EU Sales for Older vs. Newer Games')
plt.ylabel('Mean EU Sales (in millions)')
plt.xlabel('Game Category')

# Add horizontal grid lines for better readability
plt.grid(axis='y', linestyle='--', alpha=0.7)

plt.show()

# Get the counts of each unique developer and select the top 3
top_3_developers = df['Developer'].value_counts().head(3)

print(top_3_developers)

import matplotlib.pyplot as plt
import seaborn as sns

top_3_developers = pd.Series([204, 172, 167], index=['Ubisoft', 'EA Sports', 'EA Canada'])

plt.figure(figsize=(10, 6))

sns.barplot(x=top_3_developers.values, y=top_3_developers.index, palette='viridis')

for i, value in enumerate(top_3_developers.values):
    plt.text(value + 1, i, f'{value}', va='center')  # Adding a small offset (+1) for better visibility

plt.xlim(min(top_3_developers.values) - 5, max(top_3_developers.values) + 5)

plt.title('Top 3 Most Common Game Developers')
plt.xlabel('Number of Games Developed')
plt.ylabel('Developer')

plt.show()

# Aggregate sales by genre and region
genre_region_sales = df.groupby('Genre')[['NA_Sales', 'EU_Sales', 'JP_Sales', 'Other_Sales']].sum()

# Visualize genre popularity in different regions with a bar chart
genre_region_sales.plot(kind='bar', figsize=(14, 8), title='Genre Popularity by Region')
plt.ylabel('Sales (in millions)')
plt.show()

# Scatter plot for Critic Scores vs Global Sales
sns.regplot(x='Critic_Score', y='Global_Sales', data=df, scatter_kws={'alpha':0.3})
plt.title('Critic Score vs Global Sales')
plt.show()

# Scatter plot for User Scores vs Global Sales
sns.regplot(x='User_Score', y='Global_Sales', data=df, scatter_kws={'alpha':0.3})
plt.title('User Score vs Global Sales')
plt.show()

regions = ['NA_Sales', 'EU_Sales', 'JP_Sales', 'Other_Sales']

for region in regions:
    plt.figure(figsize=(10, 6))
    sns.scatterplot(x='Weighted_Rating_Score', y=region, data=genre_analysis, hue='Genre', s=100)
    plt.title(f'Genre Weighted Rating Score vs. {region}')
    plt.xlabel('Average Weighted Rating Score')
    plt.ylabel(f'Total Sales in {region} (in millions)')
    plt.legend(title='Genre', bbox_to_anchor=(1.05, 1), loc='upper left')
    plt.show()

for region in regions:
    correlation = genre_analysis['Weighted_Rating_Score'].corr(genre_analysis[region])
    print(f'Correlation between Weighted Rating Score and {region}: {correlation:.2f}')

import matplotlib.pyplot as plt

# Plotting platform sales in Japan
plt.figure(figsize=(12, 8))
platform_sales.sort_values('JP_Sales', ascending=False).plot(x='Platform', y='JP_Sales', kind='bar', color='skyblue')
plt.title('Game Sales by Platform in Japan')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(rotation=45)
plt.show()

# Plotting platform sales in North America
plt.figure(figsize=(12, 8))
platform_sales.sort_values('NA_Sales', ascending=False).plot(x='Platform', y='NA_Sales', kind='bar', color='orange')
plt.title('Game Sales by Platform in North America')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(rotation=45)
plt.show()

# Plotting platform sales in Europe
plt.figure(figsize=(12, 8))
platform_sales.sort_values('EU_Sales', ascending=False).plot(x='Platform', y='EU_Sales', kind='bar', color='green')
plt.title('Game Sales by Platform in Europe')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(rotation=45)
plt.show()

# Print top platforms in Japan
print("Top Platforms in Japan:")
print(platform_sales[['Platform', 'JP_Sales']].sort_values('JP_Sales', ascending=False).head())

# Print top platforms in North America
print("\nTop Platforms in North America:")
print(platform_sales[['Platform', 'NA_Sales']].sort_values('NA_Sales', ascending=False).head())

# Print top platforms in Europe
print("\nTop Platforms in Europe:")
print(platform_sales[['Platform', 'EU_Sales']].sort_values('EU_Sales', ascending=False).head())

# Top platforms in Japan excluding the bottom 8
top_platforms_japan = platform_sales.sort_values('JP_Sales', ascending=False).head(len(platform_sales) - 8)

plt.figure(figsize=(12, 8))
top_platforms_japan.plot(x='Platform', y='JP_Sales', kind='bar', color='skyblue')
plt.title('Top Game Sales by Platform in Japan')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(rotation=45)
plt.show()

import matplotlib.pyplot as plt
import seaborn as sns

# Ensure top platforms are sorted by sales in descending order
top_platforms_na = top_platforms_na.sort_values('NA_Sales', ascending=False)

plt.figure(figsize=(12, 8))
# Using seaborn's barplot for enhanced visuals and automatic sorting
sns.barplot(x='NA_Sales', y='Platform', data=top_platforms_na, palette='viridis')

plt.title('Top Game Sales by Platform in North America', fontsize=16)
plt.xlabel('Total Sales (in millions)', fontsize=14)
plt.ylabel('Platform', fontsize=14)
plt.xticks(fontsize=12)
plt.yticks(fontsize=12)

# Adding data labels
for index, value in enumerate(top_platforms_na['NA_Sales']):
    plt.text(value, index, f'{value:.2f}', va='center', fontsize=10)  

# Lighten gridlines and remove top and right spines
plt.grid(axis='x', linestyle='--', alpha=0.6)
sns.despine(top=True, right=True)

plt.tight_layout()
plt.show()

# Top platforms in Europe excluding the bottom 8
top_platforms_europe = platform_sales.sort_values('EU_Sales', ascending=False).head(len(platform_sales) - 8)

plt.figure(figsize=(12, 8))
top_platforms_europe.plot(x='Platform', y='EU_Sales', kind='bar', color='green')
plt.title('Top Game Sales by Platform in Europe')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(rotation=45)
plt.show()

fig, ax = plt.subplots(1, 3, figsize=(18, 6))

# Japan
top_platforms_jp.plot(ax=ax[0], x='Platform', y='JP_Sales', kind='bar', color='skyblue')
ax[0].set_title('Top Game Platforms in Japan')
ax[0].set_xlabel('Platform')
ax[0].set_ylabel('Total Sales (in millions)')
ax[0].tick_params(axis='x', rotation=45)

# North America
top_platforms_na.plot(ax=ax[1], x='Platform', y='NA_Sales', kind='bar', color='orange')
ax[1].set_title('Top Game Platforms in North America')
ax[1].set_xlabel('Platform')
ax[1].set_ylabel('Total Sales (in millions)')
ax[1].tick_params(axis='x', rotation=45)

# Europe
top_platforms_eu.plot(ax=ax[2], x='Platform', y='EU_Sales', kind='bar', color='green')
ax[2].set_title('Top Game Platforms in Europe')
ax[2].set_xlabel('Platform')
ax[2].set_ylabel('Total Sales (in millions)')
ax[2].tick_params(axis='x', rotation=45)

plt.tight_layout()
plt.show()

import pandas as pd


top_platforms_jp = platform_sales.sort_values('JP_Sales', ascending=False).head(-11)
top_platforms_na = platform_sales.sort_values('NA_Sales', ascending=False).head(-11)
top_platforms_eu = platform_sales.sort_values('EU_Sales', ascending=False).head(-11)

# Merging the top platforms DataFrames on 'Platform'
combined_platforms = pd.merge(top_platforms_jp, top_platforms_na, on='Platform', how='outer', suffixes=('_jp', '_na'))
combined_platforms = pd.merge(combined_platforms, top_platforms_eu, on='Platform', how='outer')

# Fill NaN values with 0 if any platform is not present in all regions
combined_platforms.fillna(0, inplace=True)

plt.figure(figsize=(14, 8))
width = 0.25  
positions = np.arange(len(combined_platforms['Platform']))

plt.bar(positions - width, combined_platforms['JP_Sales'], width, label='Japan', color='skyblue')
plt.bar(positions, combined_platforms['NA_Sales'], width, label='North America', color='orange')
plt.bar(positions + width, combined_platforms['EU_Sales'], width, label='Europe', color='green')

plt.title('Comparative Game Sales by Top Platforms Across Regions')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(positions, combined_platforms['Platform'], rotation=45)
plt.legend()

plt.show()

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

excluded_platforms = ['DC', 'GEN', '2600', 'SAT']
filtered_combined_platforms = combined_platforms[~combined_platforms['Platform'].isin(excluded_platforms)]

plt.figure(figsize=(14, 8))

width = 0.25  

positions = np.arange(len(filtered_combined_platforms['Platform']))

plt.bar(positions - width, filtered_combined_platforms['JP_Sales'], width, label='Japan', color='skyblue')
plt.bar(positions, filtered_combined_platforms['NA_Sales'], width, label='North America', color='orange')
plt.bar(positions + width, filtered_combined_platforms['EU_Sales'], width, label='Europe', color='green')

plt.title('Comparative Game Sales by Top Platforms Across Regions')
plt.xlabel('Platform')
plt.ylabel('Total Sales (in millions)')
plt.xticks(positions, filtered_combined_platforms['Platform'], rotation=45)
plt.legend()

plt.show()

import seaborn as sns
import matplotlib.pyplot as plt

def plot_genre_preferences(region_top_platforms, region_sales_col, region_name):
    # Filter data for the top platforms in the region
    df_top_platforms = df[df['Platform'].isin(region_top_platforms)]

    # Aggregate sales by platform and genre
    genre_sales = df_top_platforms.groupby(['Platform', 'Genre'])[region_sales_col].sum().unstack().fillna(0)

    # Plot
    genre_sales.plot(kind='bar', stacked=True, figsize=(14, 8), colormap='viridis')
    plt.title(f'Genre Preferences for Top Platforms in {region_name}')
    plt.xlabel('Platform')
    plt.ylabel(f'Total Sales in {region_name} (in millions)')
    plt.legend(title='Genre', bbox_to_anchor=(1.05, 1), loc='upper left')
    plt.xticks(rotation=45)
    plt.show()

# Plot genre preferences for top platforms in Japan
plot_genre_preferences(top5_jp, 'JP_Sales', 'Japan')

def plot_ratings_impact(region_top_platforms, region_sales_col, region_name):
    # Filter data for top platforms
    df_top_platforms = df[df['Platform'].isin(region_top_platforms)]

    # Plot
    plt.figure(figsize=(14, 8))
    sns.scatterplot(data=df_top_platforms, x='Critic_Score', y=region_sales_col, hue='Platform', style='Platform', alpha=0.6)
    plt.title(f'Impact of Critic Scores on Sales in {region_name}')
    plt.xlabel('Critic Score')
    plt.ylabel(f'Total Sales in {region_name} (in millions)')
    plt.legend(title='Platform', bbox_to_anchor=(1.05, 1), loc='upper left')
    plt.show()

# Plot the impact of game ratings on sales for top platforms in Japan
plot_ratings_impact(top5_jp, 'JP_Sales', 'Japan')

summary = highly_rated_and_high_sales[['Critic_Score', 'User_Score', 'Global_Sales']].describe()
print(summary)

plt.figure(figsize=(10, 6))
sns.scatterplot(data=highly_rated_and_high_sales, x='Critic_Score', y='Global_Sales', alpha=0.6)
plt.title('Critic Score vs. Global Sales for Top-Performing Games')
plt.xlabel('Critic Score')
plt.ylabel('Global Sales (in millions)')
plt.show()

fig, ax = plt.subplots(figsize=(10, 6))

top_genres.plot(kind='bar', ax=ax, color='skyblue')
ax.set_title('Top Genres Among Highly Rated and High Sales Games')
ax.set_ylabel('Number of Games')
ax.set_xlabel('Genre')

plt.xticks(rotation=45)
plt.show()

# Display least popular highest-rated games in Japan
print("Least Popular Highest-Rated Games in Japan:")
print(least_popular_highest_rated_jp[['Name', 'Critic_Score', 'JP_Sales']])

# Display least popular highest-rated games in North America
print("\nLeast Popular Highest-Rated Games in North America:")
print(least_popular_highest_rated_na[['Name', 'Critic_Score', 'NA_Sales']])

# Display least popular highest-rated games in Europe
print("\nLeast Popular Highest-Rated Games in Europe:")
print(least_popular_highest_rated_eu[['Name', 'Critic_Score', 'EU_Sales']])