#!/usr/bin/env python
# coding: utf-8

# # Student Alcohol Cunsumption
# ## by Nuttaphat Arunoprayoch

# Student life is messy and stressful, eventually it might lead to depression. However, we all know that cracking open a cold one with your lads always helps. In this analysis, I, for no particular reasons, would like to welcome you all to dive into Alcohol Consumption by students. Enjoy :)

# ## Dataset

# This dataset is provided and maintained by UCI. The data were obtained in a survey of students math secondary school.

# In[105]:


# Data Manipulation
import pandas as pd
import numpy as np
import pandas_profiling

# Data Visualization
get_ipython().run_line_magic('matplotlib', 'inline')
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns
from pywaffle import Waffle

mpl.style.use(['ggplot']) # optional: for ggplot-like style
sns.set(rc={'figure.figsize':(11.7,8.27)})


# In[140]:


# Hide Warning
import warnings
warnings.filterwarnings('ignore')


# In[5]:


# Dalc - workday alcohol consumption (numeric: from 1 - very low to 5 - very high)
# Walc - weekend alcohol consumption (numeric: from 1 - very low to 5 - very high)

df = pd.read_csv('student-mat.csv')
df.head()


# In[37]:


# Expand to see the whole profile
pandas_profiling.ProfileReport(df)


# ## Visualization

# In[34]:


# Checking normal distribution
df.hist()


# Even though some of data are not normalized, this is just for fun, why so serious? :)
# <hr>

# ### Gender and Alcohol Consumption

# In[92]:


# Gender on Alcohol Consumption
ax = df.groupby('sex')[['Dalc', 'Walc']].mean().plot.bar()
ax.set_title('Gender on Alcohol Consumption')
ax.set_ylabel('Average of Alcohol Consumption')
ax.set_xlabel('Gender')

df.groupby('sex')[['Dalc', 'Walc']].mean().head()


# As can be observed, males consumed alcohol more than female students both weekdays and weekends. But it doesn't seem significantly different.
# <br>** <b>Dalc</b> - workday alcohol consumption (numeric: from 1 - very low to 5 - very high)
# <br>** <b>Walc</b> - weekend alcohol consumption (numeric: from 1 - very low to 5 - very high)
# <hr>

# ### Parent's cohabitation status and Alcohol Consumption

# In[93]:


# Parent's cohabitation status on Alcohol Consumption

ax = df.groupby('Pstatus')[['Dalc', 'Walc']].mean().plot.bar()
ax.set_title('Parent\'s cohabitation status on Alcohol Consumption')
ax.set_ylabel('Sum of Alcohol Consumption')
ax.set_xlabel('Parent\'s cohabitation status')

df.groupby('Pstatus')[['Dalc', 'Walc']].mean().head()


# By exploring Parent's cohabitation status, it was found that students whose parents were living apart (A) had a slightly higher alcohol consumption rate than the other. However, it should be noted that (A) reprents only 10% of all students, it still shows a higher rate. Therefore, it can be concluded that students whose parents were living apart comsumed alcohol significantly higher than the students whose parents were living together.
# <hr>

# ### Guardians and Alcohol Consumption

# In[155]:


temp_data = df.groupby('guardian')['TotalAlc'].count()
df_gu_count = temp_data.to_dict()

ax = plt.figure(
    FigureClass=Waffle, 
    rows=10, 
    values=df_gu_count, 
    legend={'loc': 'upper left', 'bbox_to_anchor': (1.1, 1)}
)

plt.title('Students\' Guardians')
print(df_gu_count)


# According to the data collected, most of students lived with their mothers, followed by fathers and other consequently.

# In[149]:


df_gu = df.groupby('guardian', axis=0).mean()

colors_list = ['gold', 'yellowgreen', 'lightcoral', 'lightskyblue', 'lightgreen', 'pink']
explode_list = [0.1, 0, 0] # ratio for each continent with which to offset each wedge.

ax = df_gu['TotalAlc'].plot(kind='pie',
                            figsize=(15, 10),
                            autopct='%1.1f%%', 
                            startangle=90,    
                            shadow=True,       
                            labels=None,         # turn off labels on pie chart
                            pctdistance=1.12,    # the ratio between the center of each pie slice and the start of the text generated by autopct 
                            colors=colors_list, # add custom colors
                            explode=explode_list
                            )

ax.set_title('Alcohol Consumption based on Students\' Guardians')
ax.legend(df_gu.index.tolist())
df_gu.head()


# As can be observed from the pie chart depicted above, the average of alcohol consumption is almost identical. However, the students whose guardians are fathers had a slightly higher of alcohol consumption (34.3%)
# <hr>

# ### Alcohol Consumption with Various Activities

# In[170]:


df_abs = df.groupby('TotalAlc')[['absences', 'failures', 'studytime', 'traveltime']].mean()

ax = df_abs.plot(kind='area', 
                 alpha=0.3, # 0-1, default value a= 0.5
                 stacked=True,
                 figsize=(20, 10),
                )

ax.set_title('Alcohol Consumption with Various Activities')
ax.set_ylabel('Level of Activities')
ax.set_xlabel('Alcohol Consumption Levels (1-10)')


#