Visualizing various types of data

Lecture 5

Dr. Greg Chism

University of Arizona
INFO 511 - Fall 2024

Setup

# Import necessary libraries
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from palmerpenguins import load_penguins

# Load the penguins dataset
penguins = load_penguins()

# Set theme
sns.set_theme(style="whitegrid")

Data Visualization

Examining data visualization

Discuss the following for the visualization in the #lecture-discussions Slack Channel.

Source: Twitter

Violin plots

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins)
plt.title('Violin Plot of Body Mass by Species')
plt.show()

Multiple geoms

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins)
sns.stripplot(x="species", y="body_mass_g", data=penguins, jitter=False, color='black')
plt.title('Violin Plot with Points of Body Mass by Species')
plt.show()

Multiple geoms

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins)
sns.stripplot(x="species", y="body_mass_g", data=penguins, jitter=True, color='black')
plt.title('Violin Plot with Points of Body Mass by Species')
plt.show()

Multiple geoms + aesthetics

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins)
sns.stripplot(x="species", y="body_mass_g", data=penguins, jitter=True, hue='species')
plt.title('Violin Plot with Jittered Points and Color by Species')
plt.legend(title='Species')
plt.show()

Multiple geoms + aesthetics

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins)
sns.stripplot(x="species", y="body_mass_g", data=penguins, jitter=True, hue='species')
plt.title('Violin Plot with Jittered Points, Color by Species, and No Legend')
plt.legend(title='Species').remove()
plt.show()

Multiple geoms + aesthetics

plt.figure(figsize=(8, 6))
sns.violinplot(x="species", y="body_mass_g", data=penguins, palette='colorblind')
sns.stripplot(x="species", y="body_mass_g", data=penguins, jitter=True, hue='species', palette='colorblind')
plt.title('Violin Plot with Jittered Points, Color by Species, No Legend, and Colorblind Palette')
plt.legend(title='Species').remove()
plt.show()

Visualizing various types of data

The way data is displayed matters

What do these three plots show?

Source: #barbarplots

Visualizing penguins

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from palmerpenguins import load_penguins

penguins = load_penguins()

penguins.head()

	species	island	bill_length_mm	bill_depth_mm	flipper_length_mm	body_mass_g	sex	year
0	Adelie	Torgersen	39.1	18.7	181.0	3750.0	male	2007
1	Adelie	Torgersen	39.5	17.4	186.0	3800.0	female	2007
2	Adelie	Torgersen	40.3	18.0	195.0	3250.0	female	2007
3	Adelie	Torgersen	NaN	NaN	NaN	NaN	NaN	2007
4	Adelie	Torgersen	36.7	19.3	193.0	3450.0	female	2007

Artwork by @allison_horst.

Univariate analysis

Univariate analysis

Analyzing a single variable:

• Numerical: histogram, box plot, density plot, etc.

• Categorical: bar plot, pie chart, etc.

Histogram

plt.figure(figsize=(8, 6))
sns.histplot(penguins['body_mass_g'], bins=30)
plt.title('Histogram of Penguin Body Mass')
plt.xlabel('Body Mass (g)')
plt.ylabel('Count')
plt.show()

Boxplot

plt.figure(figsize=(8, 6))
sns.boxplot(y=penguins['body_mass_g'])
plt.title('Boxplot of Penguin Body Mass')
plt.ylabel('Body Mass (g)')
plt.show()

Density plot

plt.figure(figsize=(10, 6))
sns.kdeplot(penguins['body_mass_g'], fill=True)
plt.title('Density Plot of Penguin Body Mass')
plt.xlabel('Body Mass (g)')
plt.ylabel('Density')
plt.show()

Bivariate analysis

Bivariate analysis

Analyzing the relationship between two variables:

• Numerical + numerical: scatterplot

• Numerical + categorical: side-by-side box plots, violin plots, etc.

• Categorical + categorical: stacked bar plots

• Using an aesthetic (e.g., fill, color, shape, etc.) or facets to represent the second variable in any plot

Side-by-side box plots

plt.figure(figsize=(8, 6))
sns.boxplot(x="species", y="body_mass_g", data=penguins)
plt.title('Side-by-side Box Plots of Body Mass by Species')
plt.xlabel('Species')
plt.ylabel('Body Mass (g)')
plt.show()

Density plots

plt.figure(figsize=(8, 6))
sns.kdeplot(data=penguins, x="body_mass_g", hue="species", fill=True)
plt.title('Density Plot of Body Mass by Species')
plt.xlabel('Body Mass (g)')
plt.ylabel('Density')
plt.show()

Multivariate analysis

Bechdel Test

The Bechdel test also known as the Bechdel-Wallace test, is a measure of the representation of women in film and other fiction. The test asks whether a work features at least two female characters who have a conversation about something other than a man. Some versions of the test also require that those two female characters have names.

Load Bechdel test data

Load the Bechdel test data with pd.read_csv()

bechdel = pd.read_csv("data/bechdel.csv")

list() the .columns names of the bechdel data:

list(bechdel.columns)

['title', 'year', 'gross_2013', 'budget_2013', 'roi', 'binary', 'clean_test']

ROI by test result

What about this plot makes it difficult to evaluate how ROI varies by Bechdel test result?

plt.figure(figsize=(8, 4))
sns.stripplot(x='roi', y='clean_test', hue='binary', data=bechdel)
plt.title('ROI by Bechdel Test Result')
plt.xlabel('ROI')
plt.ylabel('Bechdel Test Result')
plt.legend(title='Binary')
plt.show()

Movies with high ROI

What are the movies with highest ROI?

high_roi_movies = bechdel[bechdel['roi'] > 400][['title', 'roi', 'budget_2013', 'gross_2013', 'year', 'clean_test']]
print(high_roi_movies)

                        title         roi  budget_2013   gross_2013  year  \
703       Paranormal Activity  671.336857       505595  339424558.0  2007   
1319  The Blair Witch Project  648.065333       839077  543776715.0  1999   
1575              El Mariachi  583.285665        11622    6778946.0  1992   

     clean_test  
703     dubious  
1319         ok  
1575    nowomen  

ROI by test result

Zoom in: What about this plot makes it difficult to evaluate how ROI varies by Bechdel test result?

plt.figure(figsize=(8, 4))
sns.boxplot(x='roi', y='clean_test', hue='binary', data=bechdel)
plt.xlim(0, 15)
plt.title('Zoomed in ROI by Bechdel Test Result')
plt.xlabel('ROI')
plt.ylabel('Bechdel Test Result')
plt.legend(title='Binary')
plt.show()

Sneak preview…

to next week’s data wrangling pipelines…

Median ROI

median_roi = bechdel['roi'].median()
print(f"Median ROI: {median_roi}")

Median ROI: 3.9051317558839456

Median ROI by test result

median_roi_by_test = bechdel.groupby('clean_test')['roi'].median().reset_index()
print(median_roi_by_test)

  clean_test       roi
0    dubious  3.795816
1        men  3.964457
2     notalk  3.688260
3    nowomen  3.265901
4         ok  4.211049

ROI by test result with median line

What does this plot say about return-on-investment on movies that pass the Bechdel test?

plt.figure(figsize=(8, 4))
sns.boxplot(x='roi', y='clean_test', hue='binary', data=bechdel)
plt.axvline(x=4.21, color='red', linestyle='--')
plt.xlim(0, 15)
plt.title('ROI by Bechdel Test Result with Median Line')
plt.xlabel('ROI')
plt.ylabel('Bechdel Test Result')
plt.legend(title='Binary')
plt.show()