%matplotlib inline
!pip install country_converter
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
import country_converter as coco
import plotly.offline as po
import plotly.graph_objs as pg

data_science_jobs = pd.read_csv('./data/ds_salaries.csv');
data_science_jobs.head()

data_science_jobs.drop('Unnamed: 0', axis=1, inplace=True)
data_science_jobs.head()

data_science_jobs.dtypes

data_science_jobs['work_year'].value_counts()

data_scientist_frame = data_science_jobs[data_science_jobs['job_title']=='Data Scientist']
data_scientist_frame.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(data_scientist_frame['experience_level'],
                         data_scientist_frame['salary_in_usd'])


# Customize the plot
ax.set(title='Experience Level & Salary for a Data Scientist',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

ds_salary_2021 = data_scientist_frame[data_scientist_frame['work_year']==2021]
ds_salary_2021.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(ds_salary_2021['experience_level'],
                         ds_salary_2021['salary_in_usd'],
                         color='green')


# Customize the plot
ax.set(title='Experience Level & Salary for a Data Scientist in 2021',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

ds_salary_2022 = data_scientist_frame[data_scientist_frame['work_year']==2022]
ds_salary_2022.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(ds_salary_2022['experience_level'],
                         ds_salary_2022['salary_in_usd'],
                         color='yellow')


# Customize the plot
ax.set(title='Experience Level & Salary for a Data Scientist in 2022',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=(10, 6))

# Plot the data in ax1
horizontal_bar = ax1.barh(ds_salary_2022['experience_level'],
                         ds_salary_2022['salary_in_usd'],
                         color='blue')


# Customize the plot in ax1
ax1.set(title='Experience Level & Salary for a Data Scientist in 2022',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

# Plot the data in ax2
horizontal_bar = ax2.barh(ds_salary_2021['experience_level'],
                         ds_salary_2021['salary_in_usd'],
                         color='green')


# Customize the plot in ax2
ax2.set(title='Experience Level & Salary for a Data Scientist in 2021',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

ml_frame = data_science_jobs[data_science_jobs['job_title']=='Machine Learning Engineer']
ml_frame.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(ml_frame['experience_level'],
                         ml_frame['salary_in_usd'])


# Customize the plot
ax.set(title='Experience Level & Salary for a Machine Learning Engineer',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

ml_salary_2021 = ml_frame[ml_frame['work_year']==2021]
ml_salary_2021.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(ml_salary_2021['experience_level'],
                         ml_salary_2021['salary_in_usd'],
                         color='green')


# Customize the plot
ax.set(title='Experience Level & Salary for a Machine Learning Engineer in 2021',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

ml_salary_2022 = ml_frame[ml_frame['work_year']==2022]
ml_salary_2022.head()

# Set the style
plt.style.use('seaborn-whitegrid')

# OO method from scratch
fig, ax = plt.subplots(figsize=(10, 6))

# Plot the data
horizontal_bar = ax.barh(ml_salary_2022['experience_level'],
                         ml_salary_2022['salary_in_usd'],
                         color='green')


# Customize the plot
ax.set(title='Experience Level & Salary for a Machine Learning Engineer in 2022',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

fig, (ax1, ax2) = plt.subplots(nrows=1, ncols=2, figsize=(15, 6))

# Plot the data in ax1
horizontal_bar = ax1.barh(ml_salary_2022['experience_level'],
                         ml_salary_2022['salary_in_usd'],
                         color='blue')


# Customize the plot in ax1
ax1.set(title='Experience Level & Salary for a Machine Learning Engineer in 2022',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

# Plot the data in ax2
horizontal_bar = ax2.barh(ml_salary_2021['experience_level'],
                         ml_salary_2021['salary_in_usd'],
                         color='green')


# Customize the plot in ax2
ax2.set(title='Experience Level & Salary for a Machine Learning Engineer in 2021',
      xlabel='Salary (USD)',
      ylabel='Experience Level',
      );

plt.figure(figsize=(10, 6))
sns.set_style('darkgrid')
sns.histplot(x=data_science_jobs['salary_in_usd'], hue=data_science_jobs['work_year'], palette='Set1', kde=True);

plt.figure(figsize=(15,5))
plt.subplot(1,2,1)
sns.set_style('darkgrid')
sns.barplot(x = names_top,y = values_roles_top,palette='magma')
plt.xticks(rotation='90')
plt.title('Top pay roles')
plt.subplot(1,2,2)
sns.set_style('whitegrid')
sns.barplot(x = names_least,y = values_roles_least,palette='crest')
plt.xticks(rotation='90')
plt.title('Least pay roles')
plt.show()

data_science_jobs.head()

converted_country = coco.convert(names=data_science_jobs.groupby('company_location').count().sort_values(by='work_year', 
                                                                             ascending=False).index, to='ISO3')
data= dict(type='choropleth',
          colorscale='magenta',
          locations= converted_country,
          z = data_science_jobs.groupby('company_location').count().sort_values(by='work_year', ascending=False)['work_year'])
layout = dict(title='Companies By Country',
             geo= dict(projection={'type':'robinson'},
                      showlakes=False))
x= pg.Figure(data = [data],
            layout = layout)
po.iplot(x)

converted_country = coco.convert(names=data_science_jobs.groupby('company_location').mean().sort_values(by='salary_in_usd', 
                                                                             ascending=False).index, to='ISO3')
data= dict(type='choropleth',
          colorscale='rdbu',
          locations= converted_country,
          z = data_science_jobs.groupby('company_location').mean().sort_values(by='salary_in_usd', ascending=False)['salary_in_usd'])
layout = dict(title='Salaries By Country',
             geo= dict(projection={'type':'robinson'},
                      showlakes=False))
x= pg.Figure(data = [data],
            layout = layout)
po.iplot(x)