Google Play ストアのスクレイピングデータを可視化し、Androidアプリ市場を把握する

apps_with_duplicates = pd.read_csv('datasets/apps.csv', index_col=0)    

apps = apps_with_duplicates.drop_duplicates()
apps.dropna(inplace=True)
apps.drop(['Genres','Content Rating', 'Last Updated', 'Current Ver', 'Android Ver'],axis = 1)   
apps.reset_index(drop=True, inplace=True)
apps.index.name = 'ID' 

# List of characters to remove
chars_to_remove = ['+', ',', 'M', '$']
# List of column names to clean
cols_to_clean = ['Installs', 'Size', 'Price']



# Loop for each column
for col in cols_to_clean:
    # Replace each character with an empty string
    for char in chars_to_remove:
        #print(col)
        #print(char)
        apps[col] = apps[col].astype(str).str.replace(char, '')
    apps[col] = pd.to_numeric(apps[col])

apps.to_csv("apps_processed.csv")

#read the cleaned data from csv
apps = pd.read_csv("apps_processed.csv")

#Create container 
apps_container = "apps_container"

# Create containerInfo
apps_containerInfo = griddb.ContainerInfo(apps_container,
                    [["ID", griddb.Type.INTEGER],
                    ["App", griddb.Type.STRING],
                    ["Category", griddb.Type.STRING],
                    ["Rating", griddb.Type.FLOAT],
                    ["Reviews", griddb.Type.FLOAT],
                    ["Size", griddb.Type.FLOAT],
                    ["Installs", griddb.Type.STRING],
                    ["Type", griddb.Type.STRING],
                    ["Price", griddb.Type.FLOAT]],
                    griddb.ContainerType.COLLECTION, True)
                    
apps_columns = gridstore.put_container(apps_containerInfo)

# Put rows
apps_columns.put_rows(apps)

# Define the container names
apps_container = "apps_container"

# Get the containers
obtained_apps = gridstore.get_container(apps_container)
    
# Fetch all rows - language_tag_container
query = obtained_apps.query("select *")
    
rs = query.fetch(False)
print(f"{apps_container} Data")
    
# Iterate and create a list
retrieved_data= []
while rs.has_next():
    data = rs.next()
    retrieved_data.append(data)

# Convert the list to a pandas data frame
apps = pd.DataFrame(retrieved_data, columns=["ID","App","Category","Rating","Reviews","Size","Installs","Type","Price"])

# Get the data frame details
print(apps)
apps.info()

# Print the total number of apps
print('Total number of apps in the dataset = ', len(apps))

Expensiveapp = apps.iloc[appdata['Price'].idxmax()]

# Print the total number of unique categories
num_categories = len(apps['Category'].unique())
print('Number of categories = ', num_categories)

avg_app_rating = apps['Rating'].mean()
print('Average app rating = ', avg_app_rating)

# Count the number of apps in each 'Category' and sort them in descending order
num_apps_in_category = apps['Category'].value_counts().sort_values(ascending = False)
num_apps_in_category.plot.bar()
plt.show()

type_installs = apps[["Installs","Type"]].groupby(by = "Type").mean()
sns.barplot(data=type_installs, x=type_installs.index , y = "Installs" )

# Select a few popular app categories
popular_app_cats = apps[apps.Category.isin(['GAME', 'FAMILY', 'PHOTOGRAPHY',
                                            'MEDICAL', 'TOOLS', 'FINANCE',
                                            'LIFESTYLE','BUSINESS'])]

# Select apps priced below $100
apps_under_100 = popular_app_cats[popular_app_cats['Price']<100]

fig, ax = plt.subplots()
fig.set_size_inches(15, 8)

# Examine price vs category with the authentic apps (apps_under_100)
ax = sns.stripplot(x='Price', y='Category', data=apps_under_100,
                   jitter=True, linewidth=1)
ax.set_title('App pricing trend across categories after filtering for outliers')
plt.show()