st8lgwlg3

October 1, 2024

[1]: import numpy as np # linear algebra

import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)

[3]: # importing usefull libraries

import warnings
import numpy as np
import pandas as pd
import seaborn as sns
import plotly.express as px
import plotly.graph_objects as go
import bokeh.plotting as bkp
from bokeh.models import ColumnDataSource
from bokeh.io import show
warnings.filterwarnings("ignore")

[5]: df = pd.read_csv("Global EV Data 2024.csv")

# data description
custom_cmap = sns.color_palette("BrBG", as_cmap=True) # Or use any other␣
↪colormap

df.head().style.background_gradient(cmap=custom_cmap, axis=1)

[5]: <pandas.io.formats.style.Styler at 0x200b2b16390>

[7]: df.describe(include='all').style.background_gradient(cmap=custom_cmap, axis=1)

[7]: <pandas.io.formats.style.Styler at 0x200b1790f80>

[9]: df

[9]: region category parameter mode \

0 Australia Historical EV stock share Cars
1 Australia Historical EV sales share Cars
2 Australia Historical EV sales Cars
3 Australia Historical EV stock Cars
4 Australia Historical EV stock Cars
… … … … …
12649 World Projection-STEPS EV sales share Cars

1
 12650 World Projection-STEPS EV stock share Cars
12651 World Projection-APS EV charging points EV
12652 World Projection-APS EV charging points EV
12653 World Projection-STEPS EV stock share Trucks

powertrain year unit value

0 EV 2011 percent 3.900000e-04
1 EV 2011 percent 6.500000e-03
2 BEV 2011 Vehicles 4.900000e+01
3 BEV 2011 Vehicles 4.900000e+01
4 BEV 2012 Vehicles 2.200000e+02
… … … … …
12649 EV 2035 percent 5.500000e+01
12650 EV 2035 percent 3.100000e+01
12651 Publicly available fast 2035 charging points 9.400000e+06
12652 Publicly available slow 2035 charging points 1.500000e+07
12653 EV 2035 percent 9.000000e+00

[12654 rows x 8 columns]

[11]: df.isna().sum()

[11]: region 0
category 0
parameter 0
mode 0
powertrain 0
year 0
unit 0
value 0
dtype: int64

[13]: df.duplicated().sum()

[13]: 0

[15]: df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 12654 entries, 0 to 12653
Data columns (total 8 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 region 12654 non-null object
1 category 12654 non-null object
2 parameter 12654 non-null object
3 mode 12654 non-null object

2
 4 powertrain 12654 non-null object
5 year 12654 non-null int64
6 unit 12654 non-null object
7 value 12654 non-null float64
dtypes: float64(1), int64(1), object(6)
memory usage: 791.0+ KB

[17]: # Convert 'year' to int and 'value' to float if they are not already
df['year'] = df['year'].astype(int)
df['value'] = df['value'].astype(float)

[19]: df['value'].describe()

[19]: count 1.265400e+04

mean 4.273742e+05
std 6.860498e+06
min 1.200000e-06
25% 2.000000e+00
50% 1.300000e+02
75% 5.500000e+03
max 4.400000e+08
Name: value, dtype: float64

[21]: fig = px.histogram(df, x='value', title='Distribution of Values')

fig.show()

[65]: fig = px.box(df, x='category', y='value', title='Comparison of Categories')

fig.show()

3
[67]: fig = px.bar(df.groupby('region').agg({'value': 'sum'}).reset_index(),␣
↪x='region', y='value', title='Value Distribution by Region')

fig.show()

[27]: correlation_matrix = df[['year', 'value']].corr()

print(correlation_matrix)

# Plotting heatmap
import seaborn as sns
import matplotlib.pyplot as plt

sns.heatmap(correlation_matrix, annot=True, cmap='coolwarm')

plt.title('Correlation Heatmap')
plt.show()

year value
year 1.000000 0.120782
value 0.120782 1.000000

4
[69]: # Distribution of EV Sales and Stock Over the Years
fig = px.line(df[df['parameter'].isin(['EV stock', 'EV sales'])],
x='year', y='value', color='parameter',
title='Distribution of EV Sales and Stock Over the Years')
fig.show()

5
[31]: from bokeh.plotting import figure, show
from bokeh.io import output_notebook

output_notebook()

# Scatter plot of EV Stock Share vs. Year

p = figure(title="EV Stock Share vs. Year", x_axis_label='Year',␣
↪y_axis_label='EV Stock Share')

p.scatter(df[df['parameter'] == 'EV stock share']['year'],

df[df['parameter'] == 'EV stock share']['value'],
size=10, color="navy", alpha=0.5)

show(p)

[71]: # Pie chart of Powertrain distribution

powertrain_dist = df.groupby('powertrain')['value'].sum()
fig = px.pie(values=powertrain_dist.values,names=powertrain_dist.index)
fig.show()

[73]: import plotly.graph_objects as go

# Filter data for EV stock and sales

ev_stock_sales = df[df['parameter'].isin(['EV stock', 'EV sales'])]

# Create line plots for both EV stock and EV sales

fig = go.Figure()

fig.add_trace(go.Scatter(x=ev_stock_sales[ev_stock_sales['parameter'] == 'EV␣
↪stock']['year'],

y=ev_stock_sales[ev_stock_sales['parameter'] == 'EV␣
↪stock']['value'],

mode='lines+markers',

6
 name='EV Stock'))

fig.add_trace(go.Scatter(x=ev_stock_sales[ev_stock_sales['parameter'] == 'EV␣
↪sales']['year'],

y=ev_stock_sales[ev_stock_sales['parameter'] == 'EV␣
↪sales']['value'],

mode='lines+markers',
name='EV Sales'))

fig.update_layout(title='Trend Analysis of EV Stock and Sales Over the Years',

xaxis_title='Year',
yaxis_title='Value',
legend_title='Parameter')

fig.show()

[75]: # Aggregate EV stock by region

ev_stock_by_region = df[df['parameter'] == 'EV stock'].
↪groupby('region')['value'].sum().reset_index()

# Bar plot for EV stock by region

fig = px.bar(ev_stock_by_region, x='region', y='value',
title='EV Stock by Region',
labels={'value': 'Total EV Stock', 'region': 'Region'},
text='value')

fig.update_traces(texttemplate='%{text:.2s}', textposition='outside')
fig.show()

7
[39]: # Aggregate data by powertrain
powertrain_dist = df.groupby('powertrain')['value'].sum().reset_index()

# Pie chart for powertrain distribution

fig = px.pie(powertrain_dist, values='value', names='powertrain',
title='Distribution of Powertrains',
labels={'value': 'Total Value'})

fig.show()

[41]: # Calculate yearly growth rate of EV stock

ev_stock = df[df['parameter'] == 'EV stock'].groupby('year')['value'].sum().
↪pct_change().reset_index()

# Line plot for yearly growth rate of EV stock

fig = px.line(ev_stock, x='year', y='value',
title='Yearly Growth Rate of EV Stock',
labels={'value': 'Growth Rate', 'year': 'Year'})

fig.show()

8
[43]: import plotly.express as px

# Filter data for a specific parameter, such as EV stock

map_data = df[df['parameter'] == 'EV stock']

# Aggregate the data by region and year

map_data_aggregated = map_data.groupby(['region', 'year'])['value'].sum().
↪reset_index()

[45]: # Create a choropleth map

fig = px.choropleth(map_data_aggregated,
locations="region",
locationmode='country names',
color="value",
hover_name="region",
animation_frame="year",
color_continuous_scale=px.colors.sequential.Plasma,
title="EV Stock by Region Over Time",
labels={'value': 'EV Stock'})

fig.update_geos(projection_type="natural earth")
fig.update_layout(geo=dict(showframe=False, showcoastlines=False))

fig.show()

9
[77]: # Filter data for selected parameters
parameters_of_interest = ['EV stock', 'EV sales', 'EV stock share']
time_series_data = df[df['parameter'].isin(parameters_of_interest)]
# Stacked area plot to show the share of each parameter over time
fig = px.area(time_series_data, x='year', y='value', color='parameter',
title='Share of EV Parameters Over Time',
labels={'value': 'Value', 'year': 'Year'})

fig.show()

[79]: # Box plot for distribution of EV stock across regions

fig = px.box(df[df['parameter'] == 'EV stock'],
x='region', y='value',
title='Distribution of EV Stock Across Regions',
labels={'value': 'EV Stock', 'region': 'Region'})

fig.show()

10
[81]: # Bubble chart for EV stock over time with region as bubble size
bubble_data = df[df['parameter'] == 'EV stock']

fig = px.scatter(bubble_data, x='year', y='value', size='value', color='region',

title='EV Stock Over Time by Region',
labels={'value': 'EV Stock', 'year': 'Year'},
hover_name='region', size_max=60)

fig.show()

[83]: # Sunburst chart for hierarchical view of categories, parameters, and regions
fig = px.sunburst(df, path=['category', 'parameter', 'region'], values='value',
title='Hierarchical View of EV Data by Category, Parameter,␣
↪and Region',

labels={'value': 'Value'})

fig.show()

11
[55]: import plotly.express as px

fig = px.scatter_3d(df, x='year', y='value', z='region',

color='region',
title='3D Scatter Plot of EV Data',
labels={'value': 'Value', 'year': 'Year', 'region':␣
↪'Region'})

fig.show()

[85]: ev_stock_total = df[df['parameter'] == 'EV stock'].groupby('year')['value'].

↪sum().reset_index()

fig = px.line(ev_stock_total, x='year', y='value',

title='Total EV Stock Over Years',
labels={'value': 'Total EV Stock', 'year': 'Year'})
fig.show()

12
[87]: ev_data = df[df['parameter'].isin(['EV stock', 'EV sales'])].groupby(['year',␣
↪'parameter'])['value'].sum().reset_index()

fig = px.bar(ev_data, x='year', y='value', color='parameter',

title='Total EV Stock and Sales Over Years',
labels={'value': 'Total', 'year': 'Year'})
fig.show()

[89]: fig = px.histogram(df[df['parameter'] == 'EV stock share'], x='value',

title='Distribution of EV Stock Share',
labels={'value': 'EV Stock Share'})
fig.show()

13
[ ]:

14