🎨 The Art of Colors in Matplotlib

Your Palette Awaits!

Imagine you’re a painter, but instead of canvas and brushes, you have data and code. Matplotlib is your art studio, and today we’re learning how to pick the perfect colors for your masterpieces!

🌈 Color Specification Methods

What is it?

Just like crayons have names (“red”, “blue”), Matplotlib lets you tell it what color you want in many different ways!

The 5 Ways to Say “I Want This Color”

1. Named Colors - Just say the name!

plt.plot(x, y, color='red')
plt.plot(x, y, color='tomato')
plt.plot(x, y, color='skyblue')

Simple! Like asking for “the red crayon.”

2. Hex Codes - The secret color language

plt.plot(x, y, color='#FF5733')
plt.plot(x, y, color='#3498DB')

It’s like a color password: #FF5733 = a nice orange!

3. RGB Tuples - Mix your own!

plt.plot(x, y, color=(1, 0, 0))    # Pure red
plt.plot(x, y, color=(0, 0.5, 1))  # Sky blue

Numbers go from 0 to 1. Think: (Red, Green, Blue)

4. RGBA Tuples - RGB + See-through magic!

plt.plot(x, y, color=(1, 0, 0, 0.5))

The 4th number is transparency (0.5 = half see-through)

5. Single Letters - Quick shortcuts

plt.plot(x, y, color='r')  # red
plt.plot(x, y, color='g')  # green
plt.plot(x, y, color='b')  # blue
plt.plot(x, y, color='k')  # black

Quick Reference

🔄 Color Cycling

The Story

Imagine drawing 5 lines. Do you want to pick 5 colors yourself? No way! Color cycling does it automatically!

How It Works

When you draw multiple lines, Matplotlib picks a different color for each one from a list. It’s like a color playlist that keeps repeating!

for i in range(4):
    plt.plot(x, y + i)
# Each line gets a new color automatically!

Customizing Your Color Playlist

# Use a different color cycle
plt.rcParams['axes.prop_cycle'] = plt.cycler(
    color=['#FF6B6B', '#4ECDC4', '#45B7D1']
)

Now your lines will go: coral → teal → blue → coral → …

graph TD
    A[Line 1] -->|Color 1| B[coral]
    C[Line 2] -->|Color 2| D[teal]
    E[Line 3] -->|Color 3| F[blue]
    G[Line 4] -->|Back to Color 1| H[coral]

🪟 Alpha Transparency

What is Alpha?

Remember tracing paper? You could see through it! Alpha makes your colors see-through like that.

The Magic Number

• Alpha = 1 → Solid, can’t see through
• Alpha = 0.5 → Half see-through
• Alpha = 0 → Invisible!

Examples

# Half transparent red
plt.scatter(x, y, alpha=0.5, color='red')

# Using RGBA tuple
plt.bar(x, y, color=(0.2, 0.6, 1.0, 0.3))

Why Use Transparency?

• Overlapping data: See both layers!
• Density: Dark areas = more points
• Beautiful effects: Soft, artistic look

graph TD
    A[Alpha = 1.0] -->|Solid| B[Cannot see behind]
    C[Alpha = 0.5] -->|Semi-transparent| D[Partially visible]
    E[Alpha = 0.0] -->|Invisible| F[Completely hidden]

🗺️ Colormap Types

The Big Idea

A colormap is like a rainbow ruler. Give it a number, and it gives you a color!

The 4 Colormap Families

1. 🌊 Sequential - Light to Dark

Perfect for: “How much?” questions

plt.imshow(data, cmap='Blues')
plt.imshow(data, cmap='viridis')

Low values → light | High values → dark

2. ↔️ Diverging - Two directions from center

Perfect for: “Above or below average?”

plt.imshow(data, cmap='RdBu')
plt.imshow(data, cmap='coolwarm')

Negative → blue | Zero → white | Positive → red

3. 🎯 Qualitative - Different categories

Perfect for: “What type?”

plt.scatter(x, y, c=categories, cmap='Set1')
plt.scatter(x, y, c=categories, cmap='tab10')

Each category gets a distinct color.

4. 🔁 Cyclic - Loops back around

Perfect for: Angles, time of day

plt.imshow(angles, cmap='twilight')
plt.imshow(hours, cmap='hsv')

End color matches start color!

graph TD
    A[Colormap Types] --> B[Sequential]
    A --> C[Diverging]
    A --> D[Qualitative]
    A --> E[Cyclic]
    B --> B1[viridis, Blues]
    C --> C1[RdBu, coolwarm]
    D --> D1[Set1, tab10]
    E --> E1[twilight, hsv]

📏 Color Normalization

The Problem

Your data goes from 0 to 1000, but colors go from 0 to 1. How do we connect them?

The Solution: Normalize!

Normalization squishes your data range into color range.

Types of Normalization

1. Linear (Default)

Maps values evenly.

from matplotlib.colors import Normalize
norm = Normalize(vmin=0, vmax=100)
plt.scatter(x, y, c=values, norm=norm)

2. Logarithmic

For data that spans huge ranges.

from matplotlib.colors import LogNorm
norm = LogNorm(vmin=1, vmax=1000)
plt.imshow(data, norm=norm)

3. Centered

For diverging data around a center point.

from matplotlib.colors import TwoSlopeNorm
norm = TwoSlopeNorm(vmin=-10, vcenter=0, vmax=50)

Visual Example

graph LR
    A[Data: 0-1000] -->|Normalize| B[Color: 0.0-1.0]
    B --> C[Pick color from colormap]

📊 Colorbar Basics

What is a Colorbar?

It’s the legend for colors! It tells you “this color means this value.”

Adding a Colorbar

img = plt.imshow(data, cmap='viridis')
plt.colorbar(img)

That’s it! One line of code!

What It Shows

• The colormap stretched out
• Numbers showing what values the colors represent

graph TD
    A[Your Plot] --> B[Add colorbar]
    B --> C[Users understand colors!]
    C --> D[Low values = dark]
    C --> E[High values = bright]

✨ Colorbar Customization

Make It Yours!

The basic colorbar works, but you can make it perfect.

Change the Label

cbar = plt.colorbar(img)
cbar.set_label('Temperature (°C)')

Change the Position

plt.colorbar(img, orientation='horizontal')
plt.colorbar(img, location='left')

Set Specific Tick Values

cbar = plt.colorbar(img)
cbar.set_ticks([0, 25, 50, 75, 100])

Shrink or Extend

plt.colorbar(img, shrink=0.8)  # 80% of height
plt.colorbar(img, extend='both')  # arrows on ends

Complete Example

fig, ax = plt.subplots()
img = ax.imshow(data, cmap='coolwarm')

cbar = plt.colorbar(img, shrink=0.8)
cbar.set_label('Score', fontsize=12)
cbar.set_ticks([0, 50, 100])

graph TD
    A[Basic Colorbar] --> B[Add Label]
    B --> C[Set Position]
    C --> D[Custom Ticks]
    D --> E[Perfect Colorbar!]

🎯 Quick Summary

🚀 You Did It!

You now know how to paint with data! Colors aren’t just pretty—they tell stories. A good color choice helps people understand your data instantly.

Remember: The best visualizations are both beautiful and clear!