Python Visualization

🎨 Python Visualization: Your Magic Paintbrush for Data

Imagine you have a magic paintbrush that can turn boring numbers into beautiful pictures. That’s exactly what Python visualization is! Just like how a painter uses brushes to create art, data scientists use Matplotlib and Seaborn to paint pictures with data.

🖌️ Part 1: Matplotlib Basics — Your First Paintbrush

Think of Matplotlib as your basic crayon box. It’s simple, reliable, and lets you draw anything you want!

What is Matplotlib?

Matplotlib is like the first set of crayons you ever got. It can:

• Draw lines (like connecting dots)
• Make bars (like building blocks)
• Create circles (like pie slices)

Your First Drawing

import matplotlib.pyplot as plt

# Your data - like counting candies
candies = [2, 5, 3, 8, 4]
days = [1, 2, 3, 4, 5]

# Draw a line connecting the dots
plt.plot(days, candies)
plt.show()

What’s happening?

• import = Getting your paintbrush ready
• plt.plot() = Drawing a line
• plt.show() = Showing your artwork!

Types of Basic Charts

graph TD
    A["Your Data"] --> B{Choose Chart}
    B --> C["Line: Shows Change"]
    B --> D["Bar: Compares Things"]
    B --> E["Scatter: Finds Patterns"]
    B --> F["Pie: Shows Parts"]

🎨 Part 2: Matplotlib Customization — Making It Pretty!

Now that you can draw, let’s add colors, titles, and decorations! It’s like adding glitter and stickers to your art project.

Adding Colors and Labels

import matplotlib.pyplot as plt

days = [1, 2, 3, 4, 5]
candies = [2, 5, 3, 8, 4]

plt.plot(days, candies,
         color='red',      # Red line!
         linewidth=3,      # Thick line
         marker='o')       # Dots at points

plt.title('My Candy Collection')  # Title
plt.xlabel('Days')                # X label
plt.ylabel('Number of Candies')   # Y label
plt.show()

The Customization Toolbox

Multiple Lines (Like Rainbow!)

plt.plot(days, candies, 'r-o',
         label='Candies')
plt.plot(days, cookies, 'b--s',
         label='Cookies')
plt.legend()  # Shows what each line means
plt.show()

Pro Tip: 'r-o' means red line with o dots!

🌊 Part 3: Seaborn Introduction — The Fancy Art Studio

If Matplotlib is crayons, Seaborn is a professional art kit! It makes beautiful charts with less work.

Why Seaborn?

• Prettier by default — Charts look amazing instantly
• Smarter — Works great with data tables
• Faster — Less code for complex charts

Getting Started

import seaborn as sns
import matplotlib.pyplot as plt

# Seaborn's built-in example data
tips = sns.load_dataset('tips')

# One line = Beautiful chart!
sns.scatterplot(data=tips,
                x='total_bill',
                y='tip')
plt.show()

Seaborn vs Matplotlib

graph TD
    A["Want a Chart?"] --> B{How fancy?}
    B -->|Simple & Custom| C["Matplotlib"]
    B -->|Beautiful & Quick| D["Seaborn"]
    C --> E["More control"]
    D --> F["Less code"]

Key Difference:

• Matplotlib: You paint every detail
• Seaborn: It paints beautifully, you just choose what

📊 Part 4: Distribution Plots — How Data Spreads Out

Imagine you’re throwing confetti. Where does most of it land? Distribution plots answer this question for data!

Histogram — Counting Buckets

import seaborn as sns
import matplotlib.pyplot as plt

# How much did people tip?
tips = sns.load_dataset('tips')
sns.histplot(data=tips, x='tip')
plt.show()

A histogram is like sorting marbles into buckets by size. It shows how many fall into each group.

KDE Plot — Smooth Mountain

sns.kdeplot(data=tips, x='tip')
plt.show()

KDE (Kernel Density Estimate) draws a smooth mountain showing where most data lives. The higher the mountain, the more data there!

Box Plot — The Five-Number Summary

sns.boxplot(data=tips, x='tip')
plt.show()

A box plot is like a quick report card:

• The box = Where most data lives (middle 50%)
• The line inside = The middle value
• The whiskers = The normal range
• The dots = Unusual outliers

graph TD
    A["Distribution Plots"] --> B["Histogram"]
    A --> C["KDE Plot"]
    A --> D["Box Plot"]
    B --> E["Shows counts in buckets"]
    C --> F["Shows smooth shape"]
    D --> G["Shows summary stats"]

🍦 Part 5: Categorical Plots — Comparing Groups

When you want to compare different categories (like ice cream flavors), these plots are your friends!

Bar Plot — Comparing Heights

sns.barplot(data=tips,
            x='day',
            y='total_bill')
plt.show()

Like measuring which day has the tallest stack of money!

Count Plot — Counting Members

sns.countplot(data=tips, x='day')
plt.show()

Simply counts how many items are in each category. Like counting how many people came each day!

Violin Plot — Shape + Summary

sns.violinplot(data=tips,
               x='day',
               y='total_bill')
plt.show()

Combines the shape of data (like KDE) with box plot info. It looks like a violin — wider where more data lives!

Strip Plot & Swarm Plot — Show Every Point

# Strip: Random scatter
sns.stripplot(data=tips,
              x='day', y='tip')

# Swarm: Organized scatter (no overlap)
sns.swarmplot(data=tips,
              x='day', y='tip')

🔥 Part 6: Seaborn Heatmaps — The Color Grid

A heatmap is like a mood ring for data! Colors show how hot (high) or cold (low) values are.

Creating a Heatmap

import seaborn as sns
import numpy as np

# Create sample data
data = np.random.rand(5, 5)

sns.heatmap(data,
            annot=True,  # Show numbers
            cmap='coolwarm')  # Color scheme
plt.show()

Correlation Heatmap — Finding Friends

tips = sns.load_dataset('tips')
correlation = tips.corr(numeric_only=True)

sns.heatmap(correlation,
            annot=True,
            cmap='RdYlGn')
plt.show()

This shows which numbers move together:

• Dark red = They go up together (best friends!)
• Dark blue = When one goes up, other goes down (opposites)
• White = No relationship (strangers)

Popular Color Maps

👯 Part 7: Pair and Regression Plots — The Big Picture

Pair Plot — See Everything at Once

sns.pairplot(tips)
plt.show()

A pair plot is like a photo album showing every possible combination of your data! It creates a grid where:

• Diagonal = Distribution of each variable
• Off-diagonal = Relationship between two variables

graph TD
    A["Pair Plot Grid"]
    A --> B["A vs B"]
    A --> C["A vs C"]
    A --> D["B vs A"]
    A --> E["B vs C"]
    A --> F["Diagonal: Each alone"]

Pair Plot with Colors

sns.pairplot(tips, hue='sex')
plt.show()

hue adds colors for different groups — like using different colored markers!

Regression Plot — Finding the Trend Line

sns.regplot(data=tips,
            x='total_bill',
            y='tip')
plt.show()

A regression plot draws the best-fit line through your scatter plot. It’s like connecting the dots to see the overall pattern!

LM Plot — Regression with Groups

sns.lmplot(data=tips,
           x='total_bill',
           y='tip',
           hue='smoker')
plt.show()

Like regplot but can show different lines for different groups!

🎯 Quick Summary

graph LR
    A["Python Visualization"] --> B["Matplotlib"]
    A --> C["Seaborn"]

    B --> D["Basics: plot, bar, scatter"]
    B --> E["Customize: colors, labels"]

    C --> F["Distribution: hist, kde, box"]
    C --> G["Categorical: bar, violin"]
    C --> H["Heatmaps: correlation"]
    C --> I["Pair & Regression"]

The Golden Rule

1. What do you want to show?

   • Change over time → Line plot
   • Compare groups → Bar/Categorical
   • Find patterns → Scatter/Regression
   • See distribution → Histogram/KDE/Box
   • See relationships → Heatmap/Pair plot

2. How fancy?

   • Simple & custom → Matplotlib
   • Beautiful & quick → Seaborn

🚀 You’re Now a Data Artist!

Remember:

• Matplotlib = Your trusty crayons (basic but flexible)
• Seaborn = Your fancy art kit (beautiful with less work)
• Together = Unstoppable visualization power!

The best way to learn is to play! Try changing colors, data, and chart types. Make mistakes. That’s how every great artist learns!

Now go paint your data masterpiece! 🎨