Ways to Set a Color's Alpha Value

Introduction

This lab is about setting a color's alpha value using the Python Matplotlib library. Alpha value is a measure of transparency, where a value of 0 means completely transparent and a value of 1 means completely opaque. We will explore two ways to set alpha value in Matplotlib: using the alpha keyword argument and using the (matplotlib_color, alpha) color format.

VM Tips

After the VM startup is done, click the top left corner to switch to the Notebook tab to access Jupyter Notebook for practice.

Sometimes, you may need to wait a few seconds for Jupyter Notebook to finish loading. The validation of operations cannot be automated because of limitations in Jupyter Notebook.

If you face issues during learning, feel free to ask Labby. Provide feedback after the session, and we will promptly resolve the problem for you.

Creating a Bar Chart with Explicit Alpha Value

In this step, we will create a bar chart using the bar method in Matplotlib. We will set the alpha value using the alpha keyword argument. All bars in the chart will have the same alpha value.

import matplotlib.pyplot as plt
import numpy as np

## Fixing random state for reproducibility.
np.random.seed(19680801)

fig, ax = plt.subplots()

x_values = [n for n in range(20)]
y_values = np.random.randn(20)

facecolors = ['green' if y > 0 else 'red' for y in y_values]
edgecolors = facecolors

ax.bar(x_values, y_values, color=facecolors, edgecolor=edgecolors, alpha=0.5)
ax.set_title("Explicit 'alpha' keyword value\nshared by all bars and edges")

plt.show()

Creating a Bar Chart with Varying Alpha Values

In this step, we will create a bar chart using the bar method in Matplotlib. We will set the alpha value using the (matplotlib_color, alpha) color format. Each bar in the chart will have a different alpha value, based on its y-value.

import matplotlib.pyplot as plt
import numpy as np

## Fixing random state for reproducibility.
np.random.seed(19680801)

fig, ax = plt.subplots()

x_values = [n for n in range(20)]
y_values = np.random.randn(20)

facecolors = ['green' if y > 0 else 'red' for y in y_values]
edgecolors = facecolors

## Normalize y values to get distinct face alpha values.
abs_y = [abs(y) for y in y_values]
face_alphas = [n / max(abs_y) for n in abs_y]
edge_alphas = [1 - alpha for alpha in face_alphas]

colors_with_alphas = list(zip(facecolors, face_alphas))
edgecolors_with_alphas = list(zip(edgecolors, edge_alphas))

ax.bar(x_values, y_values, color=colors_with_alphas,
        edgecolor=edgecolors_with_alphas)
ax.set_title('Normalized alphas for\neach bar and each edge')

plt.show()

Summary

In this lab, we learned two ways to set a color's alpha value in Matplotlib: using the alpha keyword argument and using the (matplotlib_color, alpha) color format. The alpha keyword argument sets the same alpha value for all bars in a chart, while the (matplotlib_color, alpha) color format allows us to set different alpha values for each bar based on its y-value.