Run in Google Colab

View on GitHub

Homework 6B - Visual Encoding¶

In this assignment, you will explore how to encode different data features to visual properties using the movies dataset located in Datasets/movies.json.

We will start with simpler visualizations and then move to more complex (and potentially cluttered) visualizations. Finally, you will create a focused visualization that uses only a few selected encodings.

Instructions¶

The assignment is divided into three parts:

1. Project Setup:

   • Set up your Python and Jupyter (or VSCode) environment.

   • Clone or download the repository provided in class (refer to the class notes).

2. Simple Visualizations: Create basic visualizations using one or two encodings.

3. Cluttered Visualization: Create a visualization that encodes many features simultaneously.

4. Focused Visualization: Create one or more visualizations using a few chosen encodings, and discuss your design choices.

5. Documentation:

   • Comment your code and add markdown explanations for each part of your analysis.

6. Submission:

   • Save your notebook and export as either PDF or HTML. If the visualization is not shown in the html, submit a separated version with altair html. Refer to: https://altair-viz.github.io/getting_started/starting.html#publishing-your-visualization (you can use the chart.save('chart_file.html') method).
   • Submit to Canvas.

You are free to use any libraries you prefer for visualization, but we recommend using matplotlib, seaborn, and/or altair for this assignment.

Happy coding and visualizing!

In [ ]:

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import altair as alt

# If you're running in a notebook environment that limits rows for interactive charts:
alt.data_transformers.disable_max_rows()

# Read the dataset
df = pd.read_json("../../Datasets/movies.json")
df.head()

# Let's create a new column for the ratio of US Gross to Production Budget
df['US Gross to Production Budget'] = df['US Gross'] / df['Production Budget']

import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import altair as alt # If you're running in a notebook environment that limits rows for interactive charts: alt.data_transformers.disable_max_rows() # Read the dataset df = pd.read_json("../../Datasets/movies.json") df.head() # Let's create a new column for the ratio of US Gross to Production Budget df['US Gross to Production Budget'] = df['US Gross'] / df['Production Budget']

Part 0: Simple Visualizations¶

In this section, you will build simple plots that use minimal visual encodings. This will help you get comfortable with basic mappings before adding complexity.

Task 0.1: Simple Scatterplot¶

Create a basic scatterplot that maps:

• x-axis: Production Budget

• y-axis: US Gross

Hint: You can use:

• plt.scatter() from matplotlib

• sns.scatterplot() from seaborn

• alt.Chart() from altair with mark_point()

In [ ]:

# Fill Code Here

# Fill Code Here

Task 0.2: Simple Bar Plot¶

Now, create a bar plot that shows the average IMDB Rating for each Major Genre.

Hint: You can use:

• plt.bar() from matplotlib

• sns.barplot() from seaborn

• alt.Chart() from altair with mark_bar()

In [ ]:

# Fill Code Here

# Fill Code Here

Task 0.3: Another Simple Scatterplot¶

Create a basic scatterplot that maps:

• x-axis: IMDB Rating

• y-axis: US Gross to Production Budget

Important: US Gross to Production Budget is a ratio. Thus, you should use the logarithmic scale for the y-axis to better visualize the data.

Hint: You can use:

• plt.yscale('log') from matplotlib (also works with seaborn)

• y=alt.Y('...').scale(type="log") from altair

In [ ]:

# Fill Code Here

# Fill Code Here

Task 0.4 : Another Simple Bar Plot¶

Create a bar plot that shows the average US Gross to Production Budget for each Major Genre.

Remember to use the logarithmic scale for the y-axis to better visualize the data.

What are the insights you can draw from this plot?

In [ ]:

# Fill Code Here

# Fill Code Here

Your comments and insights here:

Any interesting observations from any of the plots you created so far?

Part 1: Intermediate Visual Encodings¶

Now that you have built simple plots, let's gradually add more visual encodings.

Task 1.1: Adding Color¶

Extend one of the scatterplots from Tasks 0.1 or 0.3 by encoding:

• Color to MPAA Rating

Hint: You can use the hue parameter in seaborn or the color parameter in matplotlib.

Note: You can also use alt.Chart() with encode(color='MPAA Rating') for color encoding.

Complete the code cell below.

In [ ]:

# Fill Code Here

# Fill Code Here

Task 1.2: Adding Size¶

Further extend the scatterplot by encoding another feature:

• Size to e.g. Running Time min, IMDB Votes

This adds another dimension to the plot. Fill in the code below.

In [ ]:

# Fill Code Here

# Fill Code Here

Part 2: Cluttered Visualization¶

The next task is to create a visualization that encodes as many features as possible. Try mapping:

• Color

• Size

• Shape/Marker

• Position

• Transparency (alpha)

• Any other visual property you can think of?

Note: The goal is to demonstrate how quickly a visualization can become cluttered when too many encodings are used.

You are free to choose any combination of the above encodings. Do your worst!

Fill in the code below with your own choices for mappings.

In [ ]:

# Fill Code Here

# Fill Code Here

Part 3: Focused Visualization¶

For the final part, create more visualizations that use only a few well-chosen encodings in each. For example, you might decide to just use 3 encodings (2 positions + color, or two positions + size, or color + size + shape).

• Map color

• Map size

• Use position (X,Y)

Create a few different visualizations maintaining x and y positions, but changing the color, size, or shape encodings, to showcase different aspects of the data.

You can also use redundant encodings, e.g., using both color and shape to represent the same feature.

This will help you focus on the most important aspects of the data without overwhelming the viewer with too much information.

After creating the visualization, write a brief discussion (just a few sentences) in the markdown cell that follows. Explain:

• Why you selected these specific encodings.

• How these choices improve interpretability.

• Any trade-offs involved in reducing the number of encoded features.

In [ ]:

# Your code here

# Your code here

Discussion¶

Please write your reflections here:

• Why did you choose the specific encodings for the focused visualization?

• How do the selected visual properties (color, size, position) help in understanding the data better?

• What improvements in clarity do you observe compared to the cluttered visualization?

• What trade-offs do you notice when reducing the number of encoded features?

Use this cell to provide a brief explanation of your design decisions.