Lighthouse Labs W4D5 - Examples Notebook: Intro to EDA Instructor: Socorro E. Dominguez-Vidana

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

df = pd.read_csv('data/df_filled.csv', parse_dates = ['date'])
df.head(2)

	tavg	co	no	no2	o3	pm10	pm25	so2	date	latitude	longitude
0	18.9	0.21	9.220395	12.0	70.0	13.0	5.8	0.95	2022-08-30	52.358039	4.899700
1	18.9	0.20	1.580324	14.0	79.0	12.0	7.2	0.95	2022-08-30	52.359714	4.866208

Matplotlib

For a a simple plot, or when you want very fine-grained control over every aspect of the plot.

Example 1

plt.boxplot(df['pm25'], widths=0.2, whis=1.5)
plt.title('Matplotlib Box Plot of PM25')
plt.xlabel('PM25')
plt.show()

For a single plot the syntax is:
plt (the library) + . + method plt + . + figure(data[column])
Documentation Example

Example 2

When using matplotlib, the general workflow is consists of:

1. Using plt.subplots() to create a Figure and any number of Axes (graphs) you want.
   • You can think of a Figure as the canvas for your plot and Axes as the graphs on the canvas.
2. Using the ax. plotting methods to generate the visualizations.
3. Using the ax. customization methods to fine-tune your plots.
4. Displaying or saving the plot.

fig, axes = plt.subplots(1, 2, figsize=(14, 5))

axes[0].boxplot(df['pm25'])
axes[0].set_title('Box Plot of PM25 (default whiskers)')

axes[1].boxplot(df['pm25'], whis=2.5)
axes[1].set_title('Box Plot of PM25 (defined whisker values)')

plt.show()

Example 3

# Create Figure and Axes objects with which to do your plotting
fig, ax = plt.subplots(figsize=(5, 2))

# Plot the data using one of Matplotlib's plotting functions
ax.hist(df['o3'])

# Customize other aspects of the plot
ax.set_title('Distribution of O3 Levels')
ax.set_xlabel('O3 (µg/m³)')
ax.set_ylabel('Frecuency')

# Display the plot
plt.show()

Example 4

# Create 2 graphs, arranged in a 1row X 2column. 
# "axes" will be an array of Axes objects.
# Note: if our grid were 2D (i.e. more than 1 row and column), 
# then "axes" would be a 2D array of Axes objects.
fig, axes = plt.subplots(
    nrows=1, 
    ncols=2, 
    figsize=(14, 5)
)

# Plot 0
axes[0].hist(df['o3'])
axes[0].set_title('Distribution of O3 Levels')
axes[0].set_xlabel('O3 (µg/m³)')
axes[0].set_ylabel('Frecuency')

# Plot 1
axes[1].hist(df['o3'], bins=30, color='green')
axes[1].set_title('Distribution of O3 Levels')
axes[1].set_xlabel('O3 (µg/m³)')
axes[1].set_ylabel('Frecuency')
axes[1].set_xlim(left=0)

plt.show()

Example 5

fig, ax = plt.subplots(1, 2, figsize=(14, 5))

x_pm25 = np.linspace(df['pm25'].min(), df['pm25'].max(), 100)
x_o3 = np.linspace(df['o3'].min(), df['o3'].max(), 100)

density_pm25 = np.histogram(df['pm25'], bins=30, density=True)
ax[0].plot(density_pm25[1][:-1], density_pm25[0], label='PM25', alpha=0.5)

density_o3 = np.histogram(df['o3'], bins=30, density=True)
ax[0].plot(density_o3[1][:-1], density_o3[0], label='O3', alpha=0.5)

ax[0].set_title('KDE Plots of PM25 and O3')
ax[0].set_xlabel('Concentration (µg/m³)')
ax[0].set_ylabel('Density')
ax[0].legend()
ax[0].text(
    x=0.6,
    y=0.5,
    s='Good :)',
    fontsize=26,
    fontweight='bold',
    transform=ax[0].transAxes,
    bbox=dict(facecolor='green', alpha=0.5)
)

ax[1].hist(df['pm25'], bins=30, alpha=0.5, label='PM25', color='blue')
ax[1].hist(df['o3'], bins=30, alpha=0.5, label='O3', color='orange')
ax[1].set_title('Histograms of PM25 and O3')
ax[1].set_xlabel('Concentration (µg/m³)')
ax[1].set_ylabel('Frequency')
ax[1].legend()
ax[1].text(
    x=0.6,
    y=0.5,
    s='Bad :(',
    fontsize=26,
    fontweight='bold',
    transform=ax[1].transAxes,
    bbox=dict(facecolor='red', alpha=0.5)
)

plt.tight_layout()
plt.show()

Seaborn

We can also use the seaborn library to create these same plots (and more). Most Seaborn plotting functions take in:

• A data argument, which is your Pandas dataframe
• One or more column names from your dataframe which dictate which parts are used for plotting

Seaborn Gallery
Seaborn Tutorial
Seaborn Set Theme Documentation

Check out Seaborn's example plotting gallery here for inspiration on the different types of graphs possible with Seaborn.

Example 1

sns.boxplot(data=df, x='pm25', whis=1.5, width=0.2)
plt.title('Box Plot of PM25')
plt.show()

Example 2

Seaborn plots can be customized further using matplotlib. Any seaborn plotting function:

• Returns a matplotlib "Axes" object, which you can use just like any ordinary Matplotlib object.
• Can optionally take a Matplotlib "Axes" object as input, if you want to apply a Seaborn plot to an existing figure.

fig, axes = plt.subplots(1, 2, figsize=(14, 5))

# Default whiskers box plot
sns.boxplot(data=df, x='pm25', ax=axes[0], whis=1.5)
axes[0].set_title('Box Plot of PM25 (default whiskers)')

# Box plot with custom whiskers
sns.boxplot(data=df, x='pm25', ax=axes[1], whis=2.5)
axes[1].set_title('Box Plot of PM25 (defined whisker values)')

plt.tight_layout()
plt.show()

Example 3

fig, ax = plt.subplots(figsize=(5, 2))

sns.histplot(df['o3'], bins=30, ax=ax)

ax.set_title('Distribution of O3 Levels')
ax.set_xlabel('O3 (µg/m³)')
ax.set_ylabel('Frequency')
plt.show()

Example 4

fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(14, 5))

sns.histplot(df['o3'], bins=30, ax=axes[0])
axes[0].set_title('Distribution of O3 Levels')
axes[0].set_xlabel('O3 (µg/m³)')
axes[0].set_ylabel('Frequency')

sns.histplot(df['o3'], bins=30, color='green', ax=axes[1])
axes[1].set_title('Distribution of O3 Levels')
axes[1].set_xlabel('O3 (µg/m³)')
axes[1].set_ylabel('Frequency')
axes[1].set_xlim(left=0)

plt.tight_layout()
plt.show()

fig, ax = plt.subplots(1, 2, figsize=(14, 5))

sns.kdeplot(data=df, x='pm25', fill=True, label='PM25', ax=ax[0])
sns.kdeplot(data=df, x='o3', fill=True, label='O3', ax=ax[0])
ax[0].set_title('KDE Plots of PM25 and O3')
ax[0].legend()
ax[0].set_xlabel(None)
ax[0].text(
    x=0.6,
    y=0.5,
    s='Good :)',
    fontsize=26,
    fontweight='bold',
    transform=ax[0].transAxes,
    bbox=dict(facecolor='green', alpha=0.5)
)

sns.histplot(data=df, x='pm25', bins=30, ax=ax[1])
sns.histplot(data=df, x='o3', bins=30, ax=ax[1])
ax[1].set_title('Histograms of PM25 and O3')
ax[1].set_xlabel(None)
ax[1].text(
    x=0.6,
    y=0.5,
    s='Bad :(',
    fontsize=26,
    fontweight='bold',
    transform=ax[1].transAxes,
    bbox=dict(facecolor='red', alpha=0.5)
)

plt.show()

Plotly

Plotly's main feature over matplotlib and seaborn is its hover tool capability. Plotly can be installed:

• using pip: pip install plotly
• or conda: conda install -c plotly plotly.

JupyterLab:

For use in JupyterLab, also install the jupyterlab and ipywidgets packages:

• using pip: pip install "jupyterlab>=3" "ipywidgets>=7.6"
• or conda: conda install "jupyterlab>=3" "ipywidgets>=7.6"

You'll need jupyter-dash to add widgets such as sliders, dropdowns, and buttons to Plotly charts in JupyterLab.

• using pip: pip install jupyter-dash
• or conda: conda install -c conda-forge -c plotly jupyter-dash

Jupyter Notebook:

For use in Jupyter Notebook, install the notebook and ipywidgets packages:

• using pip: pip install "notebook>=5.3" "ipywidgets>=7.5"
• or conda: conda install "notebook>=5.3" "ipywidgets>=7.5"

import plotly.express as px
import plotly.io as pio

pio.renderers.default = 'iframe'

fig = px.box(df, y='pm25', title='Box Plot of PM25', width=600)
fig.show()

Example 2

import plotly.graph_objects as go
from plotly.subplots import make_subplots

fig = make_subplots(rows=1, cols=2, 
                    subplot_titles=('Box Plot of PM25 (default whiskers)', 
                                    'Box Plot of PM25 (defined whisker values)'))
fig.add_trace(
    go.Box(y=df['pm25'], 
           name='PM25 (default whiskers)', 
           boxpoints=False, 
           whiskerwidth=.75),
    row=1, col=1
)

fig.add_trace(
    go.Box(y=df['pm25'], 
           name='PM25 (defined whisker values)', 
           boxpoints=False, 
           whiskerwidth=1),
    row=1, col=2
)

fig.update_layout(height=500, width=800, title_text="Box Plot Comparisons")
fig.show()

Example 3

fig = go.Figure()

fig.add_trace(go.Histogram(x=df['o3'], nbinsx=30))

fig.update_layout(
    title='Distribution of O3 Levels',
    xaxis_title='O3 (µg/m³)',
    yaxis_title='Frequency',
    width=500, height=300
)
fig.show()

Example 4

fig = make_subplots(rows=1, 
                    cols=2, 
                    subplot_titles=['Distribution of O3 Levels', 
                                    'Distribution of O3 Levels (Green)'])

fig.add_trace(
    go.Histogram(x=df['o3'], nbinsx=30),
    row=1, col=1
)

fig.add_trace(
    go.Histogram(x=df['o3'], nbinsx=30, marker_color='green'),
    row=1, col=2
)

# Customize layout
fig.update_layout(
    title_text='O3 Levels Distribution Across Different Bins',
    width=800,
    height=400
)

fig.update_xaxes(range=[0, None], row=1, col=2)

fig.show()

Example 5

import plotly.express as px
import plotly.graph_objects as go

fig_kde = go.Figure()

fig_kde.add_trace(go.Histogram(
    x=df['pm25'],
    histnorm='probability density',
    name='PM25',
    opacity=0.5,
    marker=dict(color='blue'),
    showlegend=True
))

fig_kde.add_trace(go.Histogram(
    x=df['o3'],
    histnorm='probability density',
    name='O3',
    opacity=0.5,
    marker=dict(color='orange'),
    showlegend=True
))

fig_kde.update_layout(
    title='KDE Plots of PM25 and O3',
    xaxis_title='Concentration (µg/m³)',
    yaxis_title='Density',
    barmode='overlay'
)

fig_kde.show()

fig_hist = px.histogram(
    df,
    x='pm25',
    nbins=30,
    title='Histograms of PM25 and O3',
    opacity=0.5,
    color_discrete_sequence=['blue']
)


fig_hist.add_trace(px.histogram(
    df,
    x='o3',
    nbins=30,
    opacity=0.5,
    color_discrete_sequence=['orange']
).data[0])

fig_hist.update_layout(
    xaxis_title='Concentration (µg/m³)',
    yaxis_title='Frequency'
)

fig_hist.show()