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

sns.set_theme(style = "darkgrid")

data = pd.read_csv("starbucks_data.csv",sep=';')
data.head(5)

data.info()
data.describe()

# Deleting duplicates
data = data.drop_duplicates()

# Filtering out the outliers
data = data[np.abs(data["Calories"]-data["Calories"].mean())<=(3*data["Calories"].std())]
data = data[np.abs(data["Total Fat (g)"]-data["Total Fat (g)"].mean())<=(3*data["Total Fat (g)"].std())]
data = data[np.abs(data["Total Carbohydrates (g)"]-data["Total Carbohydrates (g)"].mean())<=(3*data["Total Carbohydrates (g)"].std())]
data = data[np.abs(data["Sugars (g)"]-data["Sugars (g)"].mean())<=(3*data["Sugars (g)"].std())]
data = data[np.abs(data["Protein (g)"]-data["Protein (g)"].mean())<=(3*data["Protein (g)"].std())]
data = data[np.abs(data["Caffeine (mg)"]-data["Caffeine (mg)"].mean())<=(3*data["Caffeine (mg)"].std())]

# Filling missing values with the mean
data['Calories'].fillna(data['Calories'].mean(), inplace=True)
data['Total Fat (g)'].fillna(data['Total Fat (g)'].mean(), inplace=True)
data['Total Carbohydrates (g)'].fillna(data['Total Carbohydrates (g)'].mean(), inplace=True)
data['Sugars (g)'].fillna(data['Sugars (g)'].mean(), inplace=True)
data['Protein (g)'].fillna(data['Protein (g)'].mean(), inplace=True)
data['Caffeine (mg)'].fillna(data['Caffeine (mg)'].mean(), inplace=True)

# Because Beverage_category contains only 1 value, we drop it
data = data.drop(columns=['Beverage_category'])

data

plt.figure(figsize=(12, 6))
sns.barplot(x='Beverage', y='Caffeine (mg)', data=data)
plt.title('Caffeine content in different beverages')
plt.xlabel('Beverage')
plt.ylabel('Caffeine Content (mg)')
plt.xticks(rotation=90)
plt.tight_layout()
plt.show()

plt.figure(figsize=(12, 6))
sns.barplot(x='Beverage_prep', y='Protein (g)', data=data)
plt.title('Protein content based on beverage preparation')
plt.xlabel('Beverage Preparation')
plt.ylabel('Protein Content (g)')
plt.xticks(rotation=45)
plt.tight_layout()
plt.show()

beverage_prep_counts = data['Beverage_prep'].value_counts()

plt.figure(figsize=(10, 8))
beverage_prep_counts.plot(kind='pie', autopct='%1.1f%%', startangle=140)
plt.title('Distribution of beverage preparations')
plt.ylabel('')
plt.show()

plt.figure(figsize=(10, 6))
sns.regplot(x='Calories', y='Sugars (g)', data=data, scatter_kws={'alpha':0.5})
plt.title('Calories vs sugar content')
plt.xlabel('Calories')
plt.ylabel('Sugar Content (g)')
plt.show()

correlation = data['Calories'].corr(data['Sugars (g)'])
print('Correlation between calories and sugar content: ', correlation)