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

# ## <u>Name </u> : ADVAIT GURUNATH CHAVAN
# ## <u>Contact No </u> : +91 70214 55852
# ## <u>Mail ID </u> : advaitchavan135@gmail.com 
# 
# 
# ## Oasis Infobyte Data Science Internship
# ## <u>Task 2</u> : Unemployment Analysis with Python

# <img src="demo.png"></img>

# ### 1. Importing the necessary dependencies

# In[1]:


import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from warnings import filterwarnings
filterwarnings(action='ignore')
import calendar
import datetime as dt
import plotly.io as plio
plio.templates
import plotly.express as px
import plotly.graph_objects as go
import plotly.figure_factory as ff

from IPython.display import HTML,display


# ### 2. Exploring the dataset

# In[2]:


unemp = pd.read_csv('Unemployment in India.csv')


# In[3]:


unemp


# In[4]:


unemp.info()


# In[5]:


unemp.shape


# #### From the dataset inforrmation section we can infer that there are 740 non-null rows in each of the 7 columns
# #### Also from dataset shape we can infer that there are 754 rows (null + non-null) in each of the 7 columns
# #### So there are 754 - 740 = 14 null values in each of the 7 columns of the dataset

# In[6]:


unemp.isna().sum()


# In[7]:


unemp[unemp.isnull().any(axis=1)]


# In[8]:


print(unemp[unemp.isnull().any(axis=1)].index.tolist())


# #### From above we can infer that there are null-values from row number 359 to 372 in each of the seven columns

# #### So we will remove these rows ; and make a new dataset

# In[9]:


unemp_non_null = unemp.iloc[:, :360].dropna()


# In[10]:


unemp_non_null


# In[11]:


unemp_non_null.info()


# In[12]:


unemp_non_null.shape


# In[13]:


unemp_non_null.isna().sum()


# In[14]:


unemp_non_null.Date.min(),unemp_non_null.Date.max() 


# #### Hence, we have removed all the rows from row number 359 to row number 372 that consisted of null values from all of the 7 columns

# #### Since, the null values were present in same rows from 359 to 372 in all of the 7 columns; we removed them without following the process the imputing

# ### 3. Data Transformation

# In[15]:


unemp_non_null['Date'] = pd.to_datetime(unemp_non_null['Date'],dayfirst=True)


# In[16]:


unemp_non_null['Frequency']= unemp_non_null['Frequency'].astype('category')


# In[17]:


unemp_non_null['Month'] =  unemp_non_null['Date'].dt.month


# In[18]:


unemp_non_null['Month_num'] = unemp_non_null['Month'].apply(lambda x : int(x))


# In[19]:


unemp_non_null['Month_name'] =  unemp_non_null['Month_num'].apply(lambda x: calendar.month_abbr[x])


# In[20]:


unemp_non_null['Region'] = unemp_non_null['Region'].astype('category')


# In[21]:


unemp_non_null['Year'] =  unemp_non_null['Date'].dt.year
unemp_non_null['Year_num'] = unemp_non_null['Year'].apply(lambda x : int(x))


# In[22]:


unemp_non_null.drop(columns='Year', inplace=True)


# In[23]:


##unemp_non_null.to_csv('unmeployment in India non null.csv', index=False)


# In[24]:


unemp_non_null


# ### 4. Statistical Data Exploration

# ### <p style="color: darkviolet;">Estimated Unemployment Rate (%): This represents the actual unemployment rate you want to calculate. It's the percentage of the labor force that is currently unemployed and seeking employment. </p>
# 
# 
# ### <p style="color: darkviolet;">Estimated Employed: This is the number of people who are currently employed.</p>
# 
# ### <p style="color : darkviolet;">Estimated Labour Participation Rate (%): This represents the percentage of the working-age population that is either employed or actively seeking employment. It includes both employed and unemployed individuals.</p>

# In[25]:


round(unemp_non_null[['Estimated Unemployment Rate (%)',
       'Estimated Employed', 'Estimated Labour Participation Rate (%)']].describe(),2)


# #### (A) Feature vs Region(State)

# In[26]:


feature_vs_region =  round(unemp_non_null.groupby(['Region'])[['Estimated Unemployment Rate (%)',
       'Estimated Employed', 'Estimated Labour Participation Rate (%)']].mean().reset_index(),2)
feature_vs_region
#feature_vs_region.to_csv('feature_vs_region.csv', index=False)


# In[27]:


fig = px.bar(feature_vs_region, x='Region', y='Estimated Unemployment Rate (%)',
             title='Estimated Unemployment Rate (%) vs Region', template='plotly', color_discrete_sequence=['gold'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Unemployment Rate (%)')
fig.update_traces(text=feature_vs_region['Estimated Unemployment Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# In[68]:


fig = px.bar(feature_vs_region, x='Region', y='Estimated Employed',
             title='Estimated Employed Count vs Region', template='plotly', color_discrete_sequence=['gold'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Employed Count')
fig.update_traces(text=feature_vs_region['Estimated Employed'])

fig.update_layout(height=1500, width=1000)

# Show the plot
fig.show()


# In[29]:


fig = px.bar(feature_vs_region, x='Region', y='Estimated Labour Participation Rate (%)',
             title='Estimated Labour Participation Rate (%) vs Region', template='plotly', color_discrete_sequence=['gold'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Labour Participation Rate (%)')
fig.update_traces(text=feature_vs_region['Estimated Labour Participation Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# #### (B) Feature vs Area(Urban and Rural)

# In[30]:


feature_vs_area = round(unemp_non_null.groupby(['Area'])[['Estimated Unemployment Rate (%)',
       'Estimated Employed', 'Estimated Labour Participation Rate (%)']].mean().reset_index(),2)
feature_vs_area
#feature_vs_area.to_excel('feature_vs_year.xlsx', index=False)
#feature_vs_area.to_csv('feature_vs_area.csv', index=False)


# In[31]:


fig = px.bar(feature_vs_area, x='Area', y='Estimated Unemployment Rate (%)',
             title='Estimated Unemployment Rate (%) vs Area', template='plotly')

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Unemployment Rate (%)')

fig.update_traces(text=feature_vs_area['Estimated Unemployment Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# In[32]:


fig = px.bar(feature_vs_area, x='Area', y='Estimated Employed',
             title='Estimated Employed Count vs Area', template='plotly')

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Employed')
fig.update_traces(text=feature_vs_area['Estimated Employed'], textposition='outside')

# Show the plot
fig.show()


# In[33]:


fig = px.bar(feature_vs_area, x='Area', y='Estimated Labour Participation Rate (%)',
             title='Estimated Labour Participation Rate (%) vs Area', template='plotly')

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Labour Participation Rate (%)')
fig.update_traces(text=feature_vs_area['Estimated Labour Participation Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# #### (C) Feature vs Year and Month

# In[34]:


feature_vs_year_month = round(unemp_non_null.groupby(['Year_num','Month_num', 'Month_name'])[['Estimated Unemployment Rate (%)',
       'Estimated Employed', 'Estimated Labour Participation Rate (%)']].mean().reset_index().sort_values(by=['Year_num']),2)
feature_vs_year_month


# In[35]:


feature_vs_year_month['Month_name'] = feature_vs_year_month['Month_name'].astype(str) 
feature_vs_year_month['Year_num_Month'] = feature_vs_year_month['Year_num'].astype(str) + ' - ' + feature_vs_year_month['Month_name']
feature_vs_year_month = feature_vs_year_month.sort_values(['Year_num', 'Month_name'])
feature_vs_year_month
#feature_vs_year_month.to_csv('feature_vs_year_month.csv', index=False)


# In[36]:


# Create the bar plot
fig = px.bar(feature_vs_year_month, x='Year_num_Month', y='Estimated Unemployment Rate (%)',
             title='Estimated Unemployment Rate (%) vs Year and Month', template='plotly', color_discrete_sequence=['red'])

# Set the labels for the x and y-axes
fig.update_xaxes(title_text='Year and Month')
fig.update_yaxes(title_text='Estimated Unemployment Rate (%)')

# Add values on the bars
fig.update_traces(text=feature_vs_year_month['Estimated Unemployment Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# In[37]:


# Create the bar plot
fig = px.bar(feature_vs_year_month, x='Year_num_Month', y='Estimated Employed',
             title='Estimated Employed Count vs Year and Month', template='plotly', color_discrete_sequence=['red'])

# Set the labels for the x and y-axes
fig.update_xaxes(title_text='Year and Month')
fig.update_yaxes(title_text='Estimated Employed')

# Add values on the bars
fig.update_traces(text=feature_vs_year_month['Estimated Employed'], textposition='outside')

# Show the plot
fig.show()


# In[38]:


# Create the bar plot
fig = px.bar(feature_vs_year_month, x='Year_num_Month', y='Estimated Labour Participation Rate (%)',
             title='Estimated Labour Participation Rate (%) vs Year and Month', template='plotly', color_discrete_sequence=['red'])

# Set the labels for the x and y-axes
fig.update_xaxes(title_text='Year and Month')
fig.update_yaxes(title_text='Estimated Labour Participation Rate (%)')

# Add values on the bars
fig.update_traces(text=feature_vs_year_month['Estimated Labour Participation Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# #### (D) Feature vs Year

# In[39]:


feature_vs_year =  round(unemp_non_null.groupby(['Year_num'])[['Estimated Unemployment Rate (%)',
       'Estimated Employed', 'Estimated Labour Participation Rate (%)']].mean().reset_index(),2)
feature_vs_year
#feature_vs_year.to_csv('feature_vs_year.csv', index=False)


# In[40]:


fig = px.bar(feature_vs_year, x='Year_num', y='Estimated Unemployment Rate (%)',
             title='Estimated Unemployment Rate (%) vs Year', template='plotly', color_discrete_sequence=['aqua'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Unemployment Rate (%)')
fig.update_traces(text=feature_vs_year['Estimated Unemployment Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# In[41]:


fig = px.bar(feature_vs_year, x='Year_num', y='Estimated Employed',
             title='Estimated Employed Count vs Year', template='plotly', color_discrete_sequence=['aqua'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Employed')
fig.update_traces(text=feature_vs_year['Estimated Employed'], textposition='outside')

# Show the plot
fig.show()


# In[42]:


fig = px.bar(feature_vs_year, x='Year_num', y='Estimated Labour Participation Rate (%)',
             title='Estimated Labour Participation Rate (%) vs Year', template='plotly', color_discrete_sequence=['aqua'])

# Set the labels for the y-axis
fig.update_yaxes(title_text='Estimated Labour Participation Rate (%)')
fig.update_traces(text=feature_vs_year['Estimated Labour Participation Rate (%)'], textposition='outside')

# Show the plot
fig.show()


# ### 5. Using correlation, pairplot and scatterplot to understand the relation between the features

# In[43]:


unemp_non_null.corr()


# In[44]:


sns.heatmap(unemp_non_null.corr(), annot=True)


# In[45]:


plt.figure(figsize=(20,6))
sns.pairplot(unemp_non_null)
plt.show()


# In[46]:


plt.figure(figsize=(20,20))
fig = px.scatter_matrix(unemp_non_null,template='plotly',
    dimensions=['Estimated Unemployment Rate (%)','Estimated Employed',
                'Estimated Labour Participation Rate (%)'],
    color='Region')
fig.show()


# In[47]:


plt.figure(figsize=(20,16))
fig = px.scatter_matrix(unemp_non_null,template='plotly',
    dimensions=['Estimated Unemployment Rate (%)','Estimated Employed',
                'Estimated Labour Participation Rate (%)'],
    color='Area')
fig.show()


# ### 6. Exploratory Data Analysis

# #### [A] Estimated Unemployment Rate(%)

# In[48]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Region',y='Estimated Unemployment Rate (%)',color='Region',title='Estimated Unemployment rate(%) vs State',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[49]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Area',y='Estimated Unemployment Rate (%)',color='Area',title='Estimated Unemployment rate(%) vs Area',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[50]:


df_est_unemp_rate_vs_region = unemp_non_null[['Estimated Unemployment Rate (%)','Region']]

df_est_unemp_rate_vs_region = df_est_unemp_rate_vs_region.groupby('Region').mean().reset_index()

df_est_unemp_rate_vs_region = df_est_unemp_rate_vs_region.sort_values('Estimated Unemployment Rate (%)')

fig = px.bar(df_est_unemp_rate_vs_region, x='Region',y='Estimated Unemployment Rate (%)',color='Region',
            title='Average Estimated Unemployment Rate(%) in each state',template='plotly', text='Estimated Unemployment Rate (%)', height=1000)

fig.show()


# In[51]:


unemp_2020 = unemp_non_null[unemp_non_null['Year_num'] == 2020]
unemp_2019 = unemp_non_null[unemp_non_null['Year_num'] == 2019]


# In[52]:


plt.figure(figsize=(20,30))
fig = px.bar(unemp_2020, x='Region',y='Estimated Unemployment Rate (%)',animation_frame = 'Month_name' ,color='Area',
        title='Estimated Unemployment rate across Region(States) during year 2020', height=900,template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000
fig.show()


# In[53]:


plt.figure(figsize=(20,30))
fig = px.bar(unemp_2019, x='Region',y='Estimated Unemployment Rate (%)',animation_frame = 'Month_name' ,color='Area',
        title='Estimated Unemployment rate across Region(States) during year 2019', height=700,template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000
fig.show()


# In[54]:


fig = px.sunburst(unemp_non_null.groupby(['Region'])['Estimated Unemployment Rate (%)'].mean().reset_index(),
                  path=['Region'], values='Estimated Unemployment Rate (%)',
                  color_continuous_scale='Plasma', title='Estimated Unemployment Rate (%) by Region(State)',
                  height=950, template='ggplot2',custom_data=['Estimated Unemployment Rate (%)'])

fig.update_traces(textinfo='label+value')

fig.show()


# #### <p style="color: red;">From the pieplot, avg. unemployment rate(%) bar plot and box plots we can infer the following:-</p>
# #### <p style="color: red;">The top 5 regions(states) in India having the highest unemployement rate (%) during COVID-19 lockdown are:</p>
# #### <p style="color: red;">1. Tripura = 28.35%</p>
# #### <p style="color: red;">2. Haryana = 26.28%</p>
# #### <p style="color: red;">3. Jharkhand = 20.59%</p>
# #### <p style="color: red;">4. Bihar = 18.92%</p>
# #### <p style="color: red;">5. Himachal Pradesh = 18.54%</p>

# #### [B] Estimated Employed Count

# In[55]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Region',y='Estimated Employed',color='Region',title='Estimated Employed Count vs State',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[56]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Area',y='Estimated Employed',color='Area',title='Estimated Employed Count vs Area',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[57]:


df_est_emp_vs_region = unemp_non_null[['Estimated Employed','Region']]

df_est_emp_vs_region = df_est_emp_vs_region.groupby('Region').mean().reset_index()

df_est_emp_vs_region = df_est_emp_vs_region.sort_values('Estimated Employed')

fig = px.bar(df_est_emp_vs_region, x='Region',y='Estimated Employed',color='Region',
            title='Average Estimated Employed in each state',template='plotly', text='Estimated Employed', height=1000)

fig.show()


# In[58]:


plt.figure(figsize=(20,25))
fig = px.bar(unemp_2020, x='Region',y='Estimated Employed',animation_frame = 'Month_name' ,color='Area',
            title='Estimated Employed count across Region(States) in India in 2020', height=700,template='plotly')

fig.update_layout(xaxis={'categoryorder':'total descending'})

fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000

fig.show()


# In[59]:


plt.figure(figsize=(20,25))
fig = px.bar(unemp_2019, x='Region',y='Estimated Employed',animation_frame = 'Month_name' ,color='Area',
            title='Estimated Employed count across Region(States) of India in 2019', height=700,template='plotly')

fig.update_layout(xaxis={'categoryorder':'total descending'})

fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000

fig.show()


# In[60]:


fig = px.sunburst(unemp_non_null.groupby(['Region'])['Estimated Employed'].mean().reset_index(),
                  path=['Region'], values='Estimated Employed',
                  color_continuous_scale='Plasma', title='Estimated Employed Count by Region(State)',
                  height=1050, template='ggplot2',custom_data=['Estimated Employed'])

fig.update_traces(textinfo='label+value')

fig.show()


# #### <p style="color: red;">From the pieplot, avg. employed count bar plot and box plots we can infer the following:-</p>
# #### <p style="color: red;">The top 5 regions(states) in India having the highest employed count during COVID-19 lockdown are:</p>
# #### <p style="color: red;">1. Uttar Pradesh = 28.09 Million </p>
# #### <p style="color: red;">2. Maharashtra = 19.99 Million</p>
# #### <p style="color: red;">3. West Bengal = 17.19 Million </p>
# #### <p style="color: red;">4. Bihar = 12.37 Million</p>
# #### <p style="color: red;">5. Tamil Nadu = 12.27 Million</p>

# #### [C] Estimated Labour Participation Rate (%)

# In[61]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Region',y='Estimated Labour Participation Rate (%)',color='Region',title='Estimated Labour Participation Rate (%) vs State',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[62]:


plt.figure(figsize=(20,15))
fig = px.box(unemp_non_null,x='Area',y='Estimated Labour Participation Rate (%)',color='Area',title='Estimated Labour Participation rate(%) vs Area',template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.show()


# In[63]:


df_est_lab_par_rate_vs_region = unemp_non_null[['Estimated Labour Participation Rate (%)','Region']]

df_est_lab_par_rate_vs_region = df_est_lab_par_rate_vs_region.groupby('Region').mean().reset_index()

df_est_lab_par_rate_vs_region = df_est_lab_par_rate_vs_region.sort_values('Estimated Labour Participation Rate (%)')

fig = px.bar(df_est_lab_par_rate_vs_region, x='Region',y='Estimated Labour Participation Rate (%)',color='Region',
            title='Average Estimated Labour Participation Rate (%) in each state',template='plotly', text='Estimated Labour Participation Rate (%)', height=1000)

fig.show()


# In[64]:


plt.figure(figsize=(20,30))
fig = px.bar(unemp_2020, x='Region',y='Estimated Labour Participation Rate (%)',animation_frame = 'Month_name' ,color='Area',
        title='Estimated Labour Participation Rate (%) across Region(States) during year 2020', height=900,template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000
fig.show()


# In[65]:


plt.figure(figsize=(20,30))
fig = px.bar(unemp_2019, x='Region',y='Estimated Labour Participation Rate (%)',animation_frame = 'Month_name' ,color='Area',
        title='Estimated Labour Participation Rate (%) across Region(States) during year 2019', height=700,template='plotly')
fig.update_layout(xaxis={'categoryorder':'total descending'})
fig.layout.updatemenus[0].buttons[0].args[1]["frame"]["duration"] = 2000
fig.show()


# In[66]:


fig = px.sunburst(unemp_non_null.groupby(['Region'])['Estimated Labour Participation Rate (%)'].mean().reset_index(),
                  path=['Region'], values='Estimated Labour Participation Rate (%)',
                  color_continuous_scale='Plasma', title='Estimated Labour Participation Rate (%) by Region(State)',
                  height=950, template='ggplot2',custom_data=['Estimated Labour Participation Rate (%)'])

fig.update_traces(textinfo='label+value')

fig.show()


# #### <p style="color: red;">From the pieplot, avg. labour participation rate(%) bar plot and box plots we can infer the following:-</p>
# #### <p style="color: red;">The top 5 regions(states) in India having the highest labour participation rate (%) during COVID-19 lockdown are:</p>
# #### <p style="color: red;">1. Tripura = 61.82%</p>
# #### <p style="color: red;">2. Meghalaya = 57.08%</p>
# #### <p style="color: red;">3. Telangana = 53.00%</p>
# #### <p style="color: red;">4. Gujarat = 46.10%</p>
# #### <p style="color: red;">5. Sikkim = 46.07%</p>

# In[ ]: