import numpy as np
import pandas as pd
import plotly.graph_objs as go
import matplotlib.colors
import warnings
warnings.filterwarnings("ignore")
import matplotlib.pyplot as plt
import seaborn as sns
import plotly as py
from datetime import datetime
import datetime as dt

amazon_csv = pd.read_csv('Amazon_prime_titles.csv')
amazon_csv.shape

np.random.seed(0) #seed for reproducibility
amazon_csv.sample(5)

# Operator to find null/ missing values
print('Percentage of null values in the given data set:')
for col in amazon_csv.columns:
  null_count = amazon_csv[col].isna().sum()/len(amazon_csv)*100
  if null_count > 0:
    print(f'{col}  : {round(null_count,2)}%')

# Define valid values
valid_types = {"Movie", "TV Show"}

# Find invalid values
invalid_types = amazon_csv[~amazon_csv['type'].isin(valid_types)]

# Check if any invalid values exist
if invalid_types.empty:
    print("Valid values")
else:
    print("\nInvalid 'type' values:")
    print(invalid_types['type'].unique())

# Check if there are any invalid ratings (less than 0 or greater than 10)
if ((amazon_csv['rating'] < 0.0) | (amazon_csv['rating'] > 10.0)).any():
    print("Invalid")
else:
    print("Valid range")

# Amazon Prime Video brand colors
sns.palplot(['#00A8E1', '#232F3E', '#FFFFFF', '#B4B4B4'])

plt.title("Amazon Prime Brand Palette", loc='left',  fontweight="bold",  fontsize=16, color='#4a4a4a', y=1.2)
plt.show()

# summary statistics
print("\nSummary Statistics:\n", amazon_csv.describe())

# new column 'duration_numeric'
amazon_csv['duration_numeric'] = amazon_csv['duration'].str.extract(r'(\d+)').astype(float)

#seaborn style and color palette
sns.set(style="whitegrid", rc={"font.family": "serif"})
palette = ['#00A8E1', '#232F3E', '#FFFFFF', '#B4B4B4']

#subplots for histograms
fig, axes = plt.subplots(1, 3, figsize=(18, 5))
fig.suptitle("Exploratory Data Analysis by plotting the Histogram and Box Plots", fontsize=16, fontweight='bold', family='serif')

# aesthetics
for ax in axes:
    ax.spines["top"].set_visible(False)
    ax.spines["right"].set_visible(False)
    ax.spines["left"].set_visible(False)
    ax.spines["bottom"].set_visible(False)

# Histogram for Release Year
sns.histplot(amazon_csv['release_year'], bins=30, kde=True, ax=axes[0], color=palette[0])
axes[0].set_title("Distribution of Release Years")
axes[0].set_xlabel("Year")

# Histogram for Rating
sns.histplot(amazon_csv['rating'].dropna(), bins=30, kde=True, ax=axes[1], color=palette[1])
axes[1].set_title("Distribution of Ratings")
axes[1].set_xlabel("Rating")

# Histogram for Duration
sns.histplot(amazon_csv['duration_numeric'].dropna(), bins=30, kde=True, ax=axes[2], color=palette[3])
axes[2].set_title("Distribution of Duration (Minutes)")
axes[2].set_xlabel("Duration (Minutes)")

plt.tight_layout()
plt.show()

# box plots
fig, axes = plt.subplots(1, 3, figsize=(18, 5))
for ax in axes:
    ax.spines["top"].set_visible(False)
    ax.spines["right"].set_visible(False)
    ax.spines["left"].set_visible(False)
    ax.spines["bottom"].set_visible(False)

sns.boxplot(y=amazon_csv['release_year'], ax=axes[0], color=palette[0])
axes[0].set_title("Box Plot of Release Years")

sns.boxplot(y=amazon_csv['rating'], ax=axes[1], color=palette[1])
axes[1].set_title("Box Plot of Ratings")

sns.boxplot(y=amazon_csv['duration_numeric'], ax=axes[2], color=palette[3])
axes[2].set_title("Box Plot of Duration (Minutes)")
plt.tight_layout()
plt.show()

#Drop the column date_added
amazon_csv.drop(columns=['date_added'], inplace=True)

# Replace missing values in 'country' with its mode (most frequent value)
amazon_csv['country'].fillna(amazon_csv['country'].mode()[0], inplace=True)

# Replace missing values in 'director' and 'cast' with "Not available"
amazon_csv[['director', 'cast']] = amazon_csv[['director', 'cast']].fillna("Not available")

# Drop rows where 'age group' or 'rating' are missing
amazon_csv.dropna(subset=['age_group', 'rating'], inplace=True)

# Reset index after dropping rows
amazon_csv.reset_index(drop=True, inplace=True)

# Set figure & Axes
fig, ax = plt.subplots(figsize=(15, 4), constrained_layout=True)
ax.set_ylim(-2, 1.75)
ax.set_xlim(0, 10)


# Timeline : line
ax.axhline(0, xmin=0.1, xmax=0.68, c='#B4B4B4', zorder=1)
# these values go on the numbers below
tl_dates = [
    "1994\nFounded",
    "2006\nIntroduced Prime",
    "2015\n100M Users",
    "2024\nPrime membership hits 250M"
]

tl_x = [1, 2.6, 4.3,6.8]
## these go on the numbers
tl_sub_x = [2.3,3.5,4.1,5.1,5.8]

tl_sub_times = [
    "2005","2010","2014","2016","2020"
]
# these values go on Stemplot : vertical line
tl_text = [
    "Amazon Prime Launched","First Amazon Original","Amazon adds streaming benefits", "India Launch (My birthplace)", "Prime Video Goes Global"]

# Timeline : Date Points
ax.scatter(tl_x, np.zeros(len(tl_x)), s=120, c='#4a4a4a', zorder=2)
ax.scatter(tl_x, np.zeros(len(tl_x)), s=30, c='#fafafa', zorder=3)
# Timeline : Time Points
ax.scatter(tl_sub_x, np.zeros(len(tl_sub_x)), s=50, c='#4a4a4a',zorder=4)

# Date Text
for x, date in zip(tl_x, tl_dates):
    ax.text(x, -0.55, date, ha='center',
             fontweight='bold',
            color='#4a4a4a',fontfamily='serif',fontsize=12)

# Stemplot : vertical line
levels = np.zeros(len(tl_sub_x))
levels[::2] = 0.3
levels[1::2] = -0.3

markerline, stemline, baseline = ax.stem(tl_sub_x, levels)
plt.setp(baseline, zorder=0)
plt.setp(markerline, marker=',', color='#4a4a4a')
plt.setp(stemline, color='#232F3E')


# Remove the Spine around the plot
for spine in ["left", "top", "right", "bottom"]:
    ax.spines[spine].set_visible(False)

# Ticks
ax.set_xticks([])
ax.set_yticks([])

# Title
ax.set_title("Amazon through the years", fontweight="bold",  fontsize=16,fontfamily='serif', color='#4a4a4a')
ax.text(2.4,1.57,"From Fast Shipping to Global Streaming - Popular streaming service for movies and TV shows.",  fontsize=12, fontfamily='serif',color='#4a4a4a')

# Text
for idx, x, time, txt in zip(range(1, len(tl_sub_x)+1), tl_sub_x, tl_sub_times, tl_text):
    ax.text(x, 1.3*(idx%2)-0.5, time, ha='center',
             fontweight='bold',
            color='#4a4a4a' if idx!=len(tl_sub_x)-1 else '#00A8E1', fontfamily='serif',fontsize=11)

    ax.text(x, 1.3*(idx%2)-0.6, txt, va='top', ha='center',fontfamily='serif',
        color='#4a4a4a' if idx!=len(tl_sub_x)-1 else '#00A8E1')
plt.show()

amazon_csv['count'] = 1
# Extract the first country from the 'country' column
amazon_csv['first_country'] = amazon_csv['country'].apply(lambda x: x.split(",")[0])

# Shorten country names
amazon_csv['first_country'].replace({'United States': 'USA','United Kingdom': 'UK','South Korea': 'S. Korea'}, inplace=True)

# Group by country and get total count
df_country = amazon_csv.groupby('first_country')['count'].sum().sort_values(ascending=False)

# Load the ISO code CSV data
iso_df = pd.read_csv('country_iso_codes.csv')


# Merge the country counts with the ISO codes
merged_df = pd.merge(countriess, iso_df, left_index=True, right_on='Country')


# Define Amazon Prime themed color scale for specific ranges for better visibility
custom_colorscale = [
    (0.0, '#00A8E1'),  # Highest values (Prime Blue)
    (0.1, '#0073CF'),  # 700-800 range (Deep Blue)
    (0.2, '#005BB5'),  # 600-700 range (Darker Blue)
    (0.3, '#004494'),  # 500-600 range (Navy Blue)
    (0.4, '#002E73'),  # 400-500 range (Deep Navy)
    (0.5, '#001A52'),  # 300-400 range (Dark Indigo)
    (0.6, '#232F3E'),  # 100-300 range (Amazon Dark Blue)
    (1.0, '#0F1111')   # 1-100 range (Almost Black)
]


# Create the choropleth map
fig = go.Figure(data=go.Choropleth(
    locations=merged_df['ISO Code'],  # Country ISO codes
    z=merged_df['count'],  # Data values
    text=merged_df['Country'],  # Hover text
    colorscale=custom_colorscale,
    marker_line_color='darkgray',
    marker_line_width=0.6,
    colorbar_title='Count',
))

# Update layout
fig.update_layout(
    title={
        'text': "TV Shows / Movies origin country available on Amazon Prime",
        'x': 0.5,
        'xanchor': 'center',
        'yanchor': 'top',
        'font': dict(
            size=30,
            family='serif',
            weight='bold'
        ),
        'pad': dict(t=20)
    },
    geo=dict(
        showframe=False,
        showcoastlines=True,
        projection_type='equirectangular'
    )
)

# Show the figure
fig.show()

# Find the ratio of type
x=amazon_csv.groupby(['type'])['type'].count()
y=len(amazon_csv)
r=((x/y)).round(2)

mf_ratio = pd.DataFrame(r).T
print(mf_ratio)

#calclated 'mf_ratio'
mf_ratio = pd.DataFrame({'Movie': [0.8], 'TV Show': [0.2]})

fig, ax = plt.subplots(figsize=(6, 1.5))

#  single horizontal stacked bar
ax.barh(0, mf_ratio['Movie'], color='#00A8E1', alpha=0.9)
ax.barh(0, mf_ratio['TV Show'], left=mf_ratio['Movie'], color='#232F3E', alpha=0.9)

# Remove axes details
ax.set_xlim(0, 1)
ax.set_xticks([])
ax.set_yticks([])
ax.spines[['top', 'left', 'right', 'bottom']].set_visible(False)

# Annotations for Movie & TV Show
ax.annotate("80%", xy=(mf_ratio['Movie'][0] / 2, 0), va='center', ha='center',  color='white', fontsize=20, fontweight='bold', fontfamily='serif')
ax.annotate("20%", xy=(mf_ratio['Movie'][0] + mf_ratio['TV Show'][0] / 2, 0), va='center', ha='center', fontsize=20, color='white', fontweight='bold', fontfamily='serif')
fig.text(1.1, 1, 'Insight', fontsize=15, fontweight='bold', fontfamily='serif')

fig.text(1.1, 0.40, '''
The graph shows that the content available
in Amazon Prime is majorly Movies with (80%)
and TV Shows  only add upto (20%).This could
be a result because of the country distribution which
we saw on the previous plot. Or Maybe because
TV Shows are something which are new to the trend than movies.
'''
         , fontsize=8, fontweight='light', fontfamily='serif')
# Creates a vertical line to seperate the insight
import matplotlib.lines as lines
l1 = lines.Line2D([1, 1], [0, 1], transform=fig.transFigure, figure=fig,color='black',lw=0.2)
fig.lines.extend([l1])

# Title
fig.text(0.125, 1.1, 'Movie & TV Show Distribution', fontsize=14, fontweight='bold', fontfamily='serif')

fig.text(0.64,0.9,"Movie", fontweight="bold", fontfamily='serif', fontsize=15, color='#00A8E1')
fig.text(0.77,0.9,"|", fontweight="bold", fontfamily='serif', fontsize=15, color='black')
fig.text(0.8,0.9,"TV Show", fontweight="bold", fontfamily='serif', fontsize=15, color='#221f1f')

plt.show()

# Filter out 'Not Available' values before counting occurrences
filtered_directors = amazon_csv[amazon_csv['director'] != "Not available"]

# Count the number of titles per director
director_counts = filtered_directors['director'].value_counts()

# Select top 10 directors
top_directors = director_counts[:6]

# Define bar colors
colors = ['#232F3E'] * 10  # Default light color
colors[:3] = ['#00A8E1'] * 3  # Highlight top 3 directors

# Create figure and axis for the bar chart
fig, ax = plt.subplots(figsize=(12, 6))

# Plot the bar chart
ax.bar(top_directors.index, top_directors.values, width=0.5, color=colors, edgecolor='darkgray', linewidth=0.6)

# Get the maximum value for dynamic annotation positioning
max_value = top_directors.max()

# Add value labels above the bars
for i, director in enumerate(top_directors.index):
    ax.annotate(f"{top_directors[director]} Content",
                xy=(i, top_directors[director] + max_value * 0.05),  # 5% above the bar
                ha='center', va='bottom', fontweight='light', fontfamily='serif')

# Remove top, left, and right borders
for side in ['top', 'left', 'right']:
    ax.spines[side].set_visible(False)

# Set x-axis labels
ax.set_xticks(range(len(top_directors.index)))  # Correct tick positioning
ax.set_xticklabels(top_directors.index, fontfamily='serif', rotation=0, ha='center')
ax.set_yticklabels('', fontfamily='serif', rotation=0, ha='center')


# Add horizontal grid lines for better readability
ax.grid(axis='y', linestyle='-', alpha=0.4)
ax.set_axisbelow(True)  # Place grid lines behind bars

# Add a thick bottom border
plt.axhline(y=0, color='black', linewidth=1.3, alpha=0.7)
fig.text(0.95, 0.75, 'Insight', fontsize=15, fontweight='bold', fontfamily='serif')

fig.text(0.95, 0.45, '''
I have chosen only Top 6 for better visualization
because majority of the directors have single
figure content under their name.
Director, Mark Knight has the highest
content of Movies/ TV show with 104 counts
that is hosted on Amazon Prime. Followed
by Cannis Holder (60), Moonbug E(36).

Its interesting although India , South Korea are one of the top
producers, the directors dont hold a Monopoly.
'''
         , fontsize=8.5, fontweight='light', fontfamily='serif')

# Remove tick marks
ax.tick_params(axis='both', which='major', labelsize=12, length=0)
fig.text(0.65,0.75,"Top 3", fontweight="bold", fontfamily='serif', fontsize=15, color='#00A8E1')
fig.text(0.71,0.75,"|", fontweight="bold", fontfamily='serif', fontsize=15, color='black')
fig.text(0.73,0.75,"Rest 3", fontweight="bold", fontfamily='serif', fontsize=15, color='#221f1f')
# Title styling
plt.title('Top 6 Directors on Amazon Prime', fontsize=16, fontweight='bold', fontfamily='serif')

# Display the plot
plt.show()

# To create a layout with length and width defined
fig, ax = plt.subplots(figsize=(12, 6))
color = ["#00A8E1", "#221f1f"]

# Annotation
for i, mtv in enumerate(amazon_csv['type'].value_counts().index):
    mtv_rel = amazon_csv[amazon_csv['type']==mtv]['release_year'].value_counts().sort_index()
    ax.plot(mtv_rel.index, mtv_rel, color=color[i], label=mtv)
    ax.fill_between(mtv_rel.index, 0, mtv_rel, color=color[i], alpha=0.9)

ax.yaxis.tick_right()
# Creates a horizontal line below in x axis
ax.axhline(y = 0, color = 'black', linewidth = 1.3, alpha = .7)
# Limits for the y axis ticks
ax.set_ylim(0, 1000)
#To remove the box around
for s in ['top', 'right','bottom','left']:
    ax.spines[s].set_visible(False)

ax.grid(False)
plt.xticks(fontfamily='serif')
plt.yticks(fontfamily='serif')
# Insights
fig.text(0.30, 0.85, 'Chronological availability of Movies & TV Shows', fontsize=16, fontweight='bold', fontfamily='serif')
fig.text(0.13, 0.75, 'Insight', fontsize=15, fontweight='bold', fontfamily='serif')
fig.text(0.13, 0.55,
'''It shows that majority of the content on Amazon Prime
is new. We can see a steep curve in the range of 2000-2020.
Definitely this will give us an insight of the age group
of the consumers.
Could this mean contents are mainly focused on youth or teens?
'''

, fontsize=8.5, fontweight='light', fontfamily='serif')
# Legends
fig.text(0.13,0.2,"Movie", fontweight="bold", fontfamily='serif', fontsize=15, color='#00A8E1')
fig.text(0.19,0.2,"|", fontweight="bold", fontfamily='serif', fontsize=15, color='black')
fig.text(0.2,0.2,"TV Show", fontweight="bold", fontfamily='serif', fontsize=15, color='#221f1f')
# removes the ticks
ax.tick_params(axis=u'both', which=u'both',length=0)
# Display the grapgh
plt.show()

# Get the unique countries
unique_countries = set()
# Get the first country if multiple are listed
for entry in amazon_csv['country'].dropna():
    first_country = entry.split(', ')[0]
    unique_countries.add(first_country)

# Filter the dataset
amazon_csv['first_country'] = amazon_csv['country'].apply(lambda x: x.split(', ')[0] if isinstance(x, str) else x)

# Get the top 10 countries
top_countries = amazon_csv['first_country'].value_counts().head(10).index
# Creating a dictionary for clubbing
ratings_ages = {
    'TV-Y': 'Kids', 'TV-G': 'Kids', 'G': 'Kids', 'ALL': 'Kids', 'ALL_AGES': 'Kids',
    'TV-Y7': 'Older Kids', '7+': 'Older Kids', 'TV-PG': 'Older Kids', 'PG': 'Older Kids',
    'PG-13': 'Teens', 'TV-14': 'Teens', '13+': 'Teens', '16+': 'Teens', '16': 'Teens', 'AGES_16_': 'Teens',
    '18+': 'Adults', 'R': 'Adults', 'TV-MA': 'Adults', 'NC-17': 'Adults', 'NR': 'Adults', 'TV-NR': 'Adults',
    'UNRATED': 'Adults', 'AGES_18_': 'Adults', 'NOT_RATE': 'Adults'
}

# Create 'target_age_group' column *before* filtering
amazon_csv['target_age_group'] = amazon_csv['age_group'].map(ratings_ages)

# Filter data for only the top 10 countries
filtered_data = amazon_csv[amazon_csv['first_country'].isin(top_countries)]
filtered_data['first_country'].replace({'United States': 'USA',
                                     'United Kingdom': 'UK',
                                     'South Korea': 'S. Korea'}, inplace=True)
# Create a pivot table with counts of each age group per country
heatmap_data = filtered_data.pivot_table(index='target_age_group', columns='first_country', aggfunc='size', fill_value=0)

# Convert counts to percentages
heatmap_data_percentage = heatmap_data.div(heatmap_data.sum(axis=0), axis=1) * 100

# custom colormap for Amazon
cmap = matplotlib.colors.LinearSegmentedColormap.from_list("", ['#FFFFFF', '#B4B4B4', '#00A8E1'])

# Plot the heatmap
plt.figure(figsize=(12, 6))
sns.heatmap(heatmap_data_percentage, cmap=cmap, annot=True, fmt='.1f', linewidths=2.5, square=True,
            cbar=False, vmax=60, vmin=5, annot_kws={"fontsize": 12, "fontfamily": 'serif', "fontweight": 'bold', "color": 'black'})

# Annotations
for t in plt.gca().texts:
    t.set_text(t.get_text() + "%")
    t.set_fontsize(10)

# Title and labels
plt.title("Target Age Group Distribution by Country (%)", fontsize=16, fontweight='bold', fontfamily='serif', pad=20, color='#232F3E')
plt.xlabel("", fontsize=10, labelpad=10, fontfamily='serif', color='#232F3E')
plt.ylabel("", fontsize=10, labelpad=10, fontfamily='serif', color='#232F3E')

plt.xticks(fontsize=10, fontfamily='serif')
plt.yticks(fontsize=10, fontfamily='serif')

# Add  labels
plt.text(10.5, 0.2, "Insight", fontweight="bold", fontfamily='serif', fontsize=16)
plt.text(10.5, 0.9, '''Yes, we can see that Most of the countries
 target Teens and Adults for their viewer engagement.
 Italy targets 70% of younger viewers.
  ''', fontfamily='serif', fontsize=8.5)

# Show plot
plt.show()

amazon_csv['num_seasons'] = amazon_csv['duration'].str.extract(r'(\d+) Season[s]?')

###Correlation between Number of Seasons Vs Ratings.

correlation = amazon_csv['num_seasons'].corr(amazon_csv['rating'])
print(f"Correlation Coefficient: {correlation}")

# Make it numeric to extract the number of seasons
amazon_csv['num_seasons'] = pd.to_numeric(amazon_csv['num_seasons'], errors='coerce')

# Create the regplot / Scatter plot
plt.figure(figsize=(12, 6))
sns.regplot(x=amazon_csv['num_seasons'], y=amazon_csv['rating'], scatter_kws={"s": 3, "color": "#232F3E"}, line_kws={"color": "#00A8E1", "linewidth": 2})

# Set title and labels
plt.title("Number of Seasons vs Rating", fontsize=16, fontweight='bold', family='serif')
plt.xlabel("Number of Seasons", fontsize=12, fontweight='normal', family='serif')
plt.ylabel("Rating", fontsize=12, fontweight='normal', family='serif')

# Grid customization
plt.grid(True, axis='both', linestyle='-', alpha=0.4)

# Make the background color white for a cleaner look
plt.gca().set_facecolor('white')
plt.text(22, 15.5, 'Insight', fontsize=16, fontweight='bold', fontfamily='serif')
plt.text(22, 13,
'''There is a strong correlation of 0.67 to show
 that rating might have tempted the Directors to
 increase the number of seasons.
'''

, fontsize=8.5, fontweight='light', fontfamily='serif')
# Customize the ticks
plt.xticks(fontsize=8, family='serif')
plt.yticks(fontsize=8, family='serif')

# Remove the borders (spines)
for spine in plt.gca().spines.values():
    spine.set_visible(False)

# Display the plot
plt.tight_layout()
plt.show()