Data science tools
Want more content like this? Subscribe here to be notified of new releases!
Data science tools
Tutorial on how to make beautiful plots with Python
python matplotlib seaborn
By Afshine Amidi and Shervine Amidi
Motivation
The Department of Transportation publicly released a dataset that lists flights that occurred in 2015, along with specificities such as delays, flight time and other information. Our previous post detailed the best practices to manipulate data.
This tutorial aims at showing good practices to visualize data using Python's most popular libraries. The following are covered:
- plots with
matplotlib,seaborn
import matplotlib.pyplot as plt
import seaborn as sns
Temporal plots
Evolution of number of flights
We want to plot the temporal evolution of flights of major US airlines:
The code used to produce the plot is shown below.
# data (date, airline, nb_flights)
# Style
sns.set(font_scale=1)
sns.set_style("whitegrid")
# Plot
fig, ax = plt.subplots(figsize=(10,3))
sns.lineplot(data=data, x='date', y='nb_flights', hue='airline')
ax.axvline(x='2015-07-01', linestyle='--', color='grey')
# Axes
ax.xaxis.set_major_locator(mdates.DayLocator(bymonthday=1))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %d'))
plt.xticks(rotation=15)
plt.ylim(0,4000)
# Legend
handles, labels = ax.get_legend_handles_labels()
labels[0] = 'Airline'
ax.legend(handles, labels, bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0, frameon=0)
plt.suptitle('Number of flights in the US in 2015', x=0.28, y=1.05)
plt.title('Top airlines', loc='left')
ax.text(x=0.75, y=-0.245, s='Source: publicly available data from DoT',
fontsize=8, transform=ax.transAxes)
plt.xlabel('Time');plt.ylabel('Number of daily flights')
Evolution of traffic per airport
We want to plot temporal evolution of traffic in the top Hawaiian airports:
The code used to produce the plot is shown below.
# data (date, airport, nb_flights), time_periods (xmin, xmax, ymin, ymax, name, color)
# Style
sns.set(font_scale=1)
sns.set_style("whitegrid")
# Plot
fig, ax = plt.subplots(figsize=(10,3))
sns.lineplot(data=data, x='date', y='nb_flights', hue='airport')
for i in range(len(time_periods)):
ax.axvspan(time_periods.loc[i, 'xmin'], time_periods.loc[i, 'xmax'],
facecolor=time_periods.loc[i, 'color'], alpha=0.2)
# Axes
ax.xaxis.set_major_locator(mdates.DayLocator(bymonthday=1))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%b %d'))
plt.xticks(rotation=15)
plt.ylim(0,150)
# Legend
handles, labels = ax.get_legend_handles_labels()
labels[0] = 'Airport'
ax.legend(handles, labels, bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0, frameon=0)
plt.title('Top 5 Hawaii airports', loc='left')
plt.suptitle('Number of flights departing from Hawaiian airports in 2015', x=0.38, y=1.05)
ax.text(x=0.75, y=-0.245, s='Source: publicly available data from DoT',
fontsize=8, transform=ax.transAxes)
plt.xlabel('Time');plt.ylabel('Number of daily flights')
Visualizing across multiple dimensions
Delay by hour of day for top airlines
We can also plot the same data from a different angle with a heat map:
The code used to produce the plot is shown below.
# Plot
fig, ax = plt.subplots(figsize=(15,5))
sns.heatmap(data, cmap='RdBu_r', ax=ax, linecolor='black', linewidth=0.01)
# Legend
ax.invert_yaxis()
ax.set_xlabel('Hour of day')
ax.set_ylabel('')
Customizing visualization
Scatter plot of airports with delays
We want to plot:
The code used to produce the plot is shown below.
# Style
sns.set(font_scale=1)
sns.set_style("whitegrid")
# Plot
fig, ax = plt.subplots(figsize=(10,3))
sns.scatterplot(data=data, x='weather', y='perc_delay', hue='region', size='nb_flights',
sizes=(50,200))
# Axes
ax.xaxis.set_major_formatter(mtick.PercentFormatter(xmax=1, decimals=0))
ax.yaxis.set_major_formatter(mtick.PercentFormatter(xmax=1, decimals=0))
# Legend
handles, labels = ax.get_legend_handles_labels()
labels[0] = 'Region';labels[5] = 'Number of flights'
ax.legend(handles, labels, bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0, frameon=0)
plt.suptitle('Percentage of delayed flights vs likelihood of weather causing the delay',
x=0.44, y=1.05)
plt.title('Top US airports across 2015', loc='left')
ax.text(x=0.75, y=-0.21, s='Source: publicly available data from DoT',
fontsize=8, transform=ax.transAxes)
plt.xlabel('Likelihood weather is causing delay');plt.ylabel('Percentage of delayed flights')
Percentage of delayed flights with volume
We want to plot:
The code used to produce the plot is shown below.
# Style
sns.set(font_scale=1.3)
sns.set_style("whitegrid")
# Plot
fig, ax = plt.subplots(figsize=(15,6))
g1 = sns.lineplot(data=data, x='date', y='nb_flights', color='green')
plt.xlabel('Time')
plt.ylabel('Number of daily flights')
plt.ylim(0,20000)
ax2 = plt.twinx()
g2 = sns.lineplot(data=data, x='date', y='perc_delay', color='red', ax=ax2)
plt.ylabel('Percentage of delayed flights')
plt.ylim(0,1)
# Legend
plt.title('Number of flights in the US in 2015', loc='left')
ax.text(x=0.7, y=-0.13, s='Source: publicly available data from DoT',
fontsize=14, transform=ax.transAxes)
Conclusion
Matplotlib has many features, but is less intuitive than R's ggplot2.
