Matplotlib Colormaps

In data science, visualization is key to understanding the structure of your data. One powerful tool for creating visualizations is **Matplotlib**, and colormaps are an essential part of making your plots informative and visually appealing. In this tutorial, we’ll explore what colormaps are, how they work in Matplotlib, and how you can use them to create beautiful, effective visualizations.

## What is a Colormap?

A **colormap** is a mapping of data values to colors. It provides a way to represent the range of data you’re plotting by mapping numbers to colors in a gradient. For example, if you're plotting temperatures, a colormap could map lower temperatures to blue and higher temperatures to red.

In Matplotlib, colormaps are used with plots like heatmaps, contour plots, scatter plots, and more, to visually represent data in a way that makes it easy to discern patterns and insights.

### Why Use Colormaps?

- **Highlight Data Trends**: Colormaps can reveal trends, clusters, or outliers in data by associating certain values with specific colors.
- **Improve Readability**: By using an appropriate colormap, you can make your plots clearer and easier to interpret.
- **Visual Appeal**: A well-chosen colormap makes your visualizations look polished and professional.

## Types of Colormaps in Matplotlib

Matplotlib provides several categories of colormaps. The right colormap to use depends on the kind of data you're working with:

### 1. Sequential Colormaps
- Used for data that has a natural order (e.g., temperature, time, population).
- These colormaps go from light to dark (or vice versa) and are useful for representing magnitude.
  
**Example Colormaps**: `Blues`, `Oranges`, `Reds`, `Greens`, `Purples`, `Greys`.


import numpy as np
import matplotlib.pyplot as plt

# Sample Data
data = np.random.randn(30, 30)
# Create a heatmap with a sequential colormap
plt.imshow(data, cmap='Blues')
plt.colorbar()  # Adds a color scale legend
plt.title('Sequential Colormap Example')
plt.show()
### 2. Diverging Colormaps
- Best for data with a meaningful midpoint (e.g., deviations from an average or zero).
- They use two contrasting colors on either side of a central point to highlight positive and negative deviations.

**Example Colormaps**: `coolwarm`, `PiYG`, `PRGn`, `RdBu`, `RdYlBu`.


import numpy as np
import matplotlib.pyplot as plt

# Sample Data
data = np.random.randn(30, 30)
# Heatmap with a diverging colormap
plt.imshow(data, cmap='coolwarm')
plt.colorbar()
plt.title('Diverging Colormap Example')
plt.show()
### 3. Qualitative (Categorical) Colormaps
- Used for categorical data, where there is no inherent ordering.
- These colormaps use distinct, contrasting colors for different categories.

**Example Colormaps**: `Set1`, `Set2`, `Accent`, `Paired`, `Dark2`.


import numpy as np
import matplotlib.pyplot as plt

# Scatter plot with a qualitative colormap
x = np.random.randn(100)
y = np.random.randn(100)
colors = np.random.randint(0, 5, 100)  # Assigning random categories

plt.scatter(x, y, c=colors, cmap='Set1')
plt.colorbar()
plt.title('Qualitative Colormap Example')
plt.show()
### 4. Cyclic Colormaps
- Used for data that is cyclic or wraps around, such as angles, days of the week, or time of day.
- Cyclic colormaps cycle through colors such that the ends meet.

**Example Colormaps**: `twilight`, `hsv`.


import numpy as np
import matplotlib.pyplot as plt

# Sample Data
data = np.random.randn(30, 30)
# Heatmap with a cyclic colormap
plt.imshow(data, cmap='twilight')
plt.colorbar()
plt.title('Cyclic Colormap Example')
plt.show()
## Customizing Colormaps

You can also customize colormaps to fit your needs. Here are a few ways to do that:

### 1. Reverse a Colormap
To reverse a colormap, just append `_r` to the name of the colormap.


import numpy as np
import matplotlib.pyplot as plt

# Sample Data
data = np.random.randn(30, 30)
plt.imshow(data, cmap='Blues_r')
plt.colorbar()
plt.title('Reversed Colormap')
plt.show()
### 2. Using `ListedColormap` for Custom Colors
If you want to create your own set of colors, you can use `ListedColormap`.


import numpy as np
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap

# Define custom colors
colors = ['#FF0000', '#00FF00', '#0000FF']
custom_cmap = ListedColormap(colors)
# Sample Data
data = np.random.randn(30, 30)
# Apply the custom colormap
plt.imshow(data, cmap=custom_cmap)
plt.colorbar()
plt.title('Custom Colormap')
plt.show()
### 3. Subset a Colormap
You can also extract a subset of a colormap by using `mpl.colors.Colormap`.


import numpy as np
import matplotlib.pyplot as plt
import matplotlib.colors as mcolors

# Create a new colormap that uses only the first half of the 'Blues' colormap
cmap = plt.get_cmap('Blues', 256)
new_cmap = mcolors.ListedColormap(cmap(np.linspace(0, 0.5, 256)))
# Sample Data
data = np.random.randn(30, 30)
plt.imshow(data, cmap=new_cmap)
plt.colorbar()
plt.title('Subset of a Colormap')
plt.show()
## Choosing the Right Colormap

Here are a few tips for choosing the best colormap:
- **Sequential colormaps** are great when data has a natural order or progression (like heat or population density).
- **Diverging colormaps** are useful when highlighting how values deviate from a neutral or central point (e.g., temperature anomalies).
- **Qualitative colormaps** should be used for categorical or non-numeric data where the categories don’t have an inherent order (e.g., types of fruits).
- **Cyclic colormaps** are ideal for periodic data (e.g., phases of the moon, seasons, time of day).

## Conclusion

Colormaps are a powerful tool for improving your visualizations and making your data easier to understand. By choosing the right colormap, you can make patterns in your data stand out and provide viewers with clearer insights. Play around with different colormaps to see how they affect the way you interpret your data.

You may also like