import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
warnings.filterwarnings('ignore')

# Load the Social Media Content Dataset
np.random.seed(42)
n_posts = 500
views = np.random.randint(100, 100000, n_posts)
df = pd.DataFrame({
    'post_id': range(1, n_posts+1),
    'platform': np.random.choice(['YouTube','Instagram','TikTok'], n_posts),
    'date': pd.date_range('2023-01-01', periods=n_posts, freq='6h'),
    'views': views,
    'likes': (views * np.random.uniform(0.02, 0.15, n_posts)).astype(int),
    'comments': (views * np.random.uniform(0.001, 0.05, n_posts)).astype(int),
    'shares': (views * np.random.uniform(0.001, 0.03, n_posts)).astype(int)
})
print(df.shape)
print(df.head(3))

# Group by date and plot views
daily_views = df.groupby(df['date'].dt.date)['views'].sum()
plt.figure(figsize=(10,4))
plt.plot(daily_views.index, daily_views.values, marker='o')
plt.title('Total Views per Day')
plt.xlabel('Date')
plt.ylabel('Views')
plt.tight_layout()
plt.show()

# Count number of posts per platform
platform_counts = df['platform'].value_counts()
sns.barplot(x=platform_counts.index, y=platform_counts.values, palette='deep')
plt.title('Number of Posts per Platform')
plt.xlabel('Platform')
plt.ylabel('Number of Posts')
plt.show()

# Calculate engagement rate for each post and take platform averages
df['engagement_rate'] = (df['likes'] + df['comments'] + df['shares']) / df['views'] * 100
avg_engagement = df.groupby('platform')['engagement_rate'].mean()
avg_engagement.plot(kind='bar', color=['#c4302b','#e1306c','#69c9d0'])
plt.title('Average Engagement Rate by Platform (%)')
plt.ylabel('Engagement Rate (%)')
plt.xlabel('Platform')
plt.show()

# Histogram and KDE for view distributions
plt.figure(figsize=(8,4))
sns.histplot(df['views'], bins=30, kde=True, color='skyblue', edgecolor='white')
plt.title('Distribution of Views per Post')
plt.xlabel('Views')
plt.ylabel('Number of Posts')
plt.tight_layout()
plt.show()

# Compute correlations and plot heatmap
corr = df[['views','likes','comments','shares','engagement_rate']].corr()
plt.figure(figsize=(7,6))
sns.heatmap(corr, annot=True, cmap='YlGnBu', fmt='.2f')
plt.title('Correlation Between Content Metrics')
plt.show()

# Get top 10 posts by engagement rate
top_posts = df.sort_values('engagement_rate', ascending=False).head(10)
display_cols = ['post_id','platform','date','views','likes','comments','shares','engagement_rate']
print(top_posts[display_cols])

# Add weekday column and compute pivot table
df['weekday'] = df['date'].dt.day_name()
pivot = df.pivot_table(index='weekday', columns='platform', values='engagement_rate', aggfunc='mean')
weekday_order = ['Monday','Tuesday','Wednesday','Thursday','Friday','Saturday','Sunday']
pivot = pivot.reindex(weekday_order)
plt.figure(figsize=(7,5))
sns.heatmap(pivot, annot=True, cmap='coolwarm', fmt='.1f')
plt.title('Avg Engagement Rate by Platform and Weekday')
plt.xlabel('Platform')
plt.ylabel('Weekday')
plt.show()

# Rolling 7-day views trend
daily_views = df.groupby(df['date'].dt.date)['views'].sum()
rolling_views = daily_views.rolling(window=7, min_periods=1).mean()
plt.figure(figsize=(10,4))
plt.plot(daily_views.index, daily_views, label='Daily Views', alpha=0.4)
plt.plot(rolling_views.index, rolling_views, color='red', label='7-Day Rolling Average')
plt.title('Daily Views and 7-Day Trend')
plt.xlabel('Date')
plt.ylabel('Views')
plt.legend()
plt.tight_layout()
plt.show()

# Simulate missing likes and fill with zeros before engaging
df_missing = df.copy()
missing_idx = np.random.choice(df.index, size=20, replace=False)
df_missing.loc[missing_idx, 'likes'] = np.nan
df_missing['likes_filled'] = df_missing['likes'].fillna(0)
# Compute engagement rate safely with the filled values
df_missing['engagement_rate_fixed'] = (df_missing['likes_filled'] + df_missing['comments'] + df_missing['shares']) / df_missing['views'] * 100
print(df_missing[['likes','likes_filled','engagement_rate_fixed']].head(10))

# Sum engagement rates incorrectly vs correctly
sum_wrong = df[df['platform']=='YouTube']['engagement_rate'].sum()
print(f"Wrong sum of engagement rates: {sum_wrong:.2f}")
sum_right = (df[df['platform']=='YouTube'][['likes','comments','shares']].sum().sum() / df[df['platform']=='YouTube']['views'].sum()) * 100
print(f"Correct overall engagement rate: {sum_right:.2f}%")

# Compare the most viewed post to its engagement rate
max_view_idx = df['views'].idxmax()
most_viewed_post = df.loc[max_view_idx]
print(f"Post ID: {most_viewed_post['post_id']}")
print(f"Platform: {most_viewed_post['platform']}")
print(f"Views: {most_viewed_post['views']}")
print(f"Engagement Rate: {most_viewed_post['engagement_rate']:.2f}%")

# Save a summary heatmap of engagement rates to file
fig, ax = plt.subplots(figsize=(6,4))
sns.heatmap(pivot, annot=True, cmap='coolwarm', fmt='.1f', ax=ax)
plt.title('Avg Engagement Rate by Platform and Weekday')
plt.xlabel('Platform')
plt.ylabel('Weekday')
fig.tight_layout()
fig.savefig('engagement_heatmap.png')
plt.close(fig)

# Which platform and weekday combination has the highest avg engagement?
best = pivot.stack().idxmax()
best_value = pivot.stack().max()
print(f"Best time/platform: {best[1]} on {best[0]} with {best_value:.2f}% average engagement rate.")

# Save summary as text recommendation
adv_txt = f"Recommendation: For highest engagement, post on {best[1]} every {best[0]}. Max observed average engagement rate is {best_value:.2f}%."
with open('content_strategy.txt','w') as f:
    f.write(adv_txt)
print(adv_txt)