Data Visualization

Data Visualization

Current Level: Intermediate
Domain: Business Analytics / Data Science

---

Overview

Data visualization transforms data into visual representations that make insights clear and actionable. Effective data visualization follows principles of clarity, accuracy, and simplicity to help users understand complex data quickly.

Data Visualization Principles

Edward Tufte's Principles

Stephen Few's Principles

Data-Ink Ratio

Data-Ink Ratio = Data-Ink / Total Ink

Goal: Maximize ratio (close to 1)

Bad: Lots of decoration, 3D effects, shadows Good: Clean, minimal, data-focused

Chart Types and Use Cases

Comprehensive Chart Guide

Line Chart

Use for: Time series, trends

Revenue ($)
$100k ┤
 $80k ┤  ●───●───●───●───●
 $60k ┤
 $40k ┤
 $20k ┤
   $0 └────────────────────────────
        Jan  Feb  Mar  Apr  May

Best practices:

• Smooth curves for trends
• Don't connect unrelated points
• Use area charts for cumulative
• Max 5-7 lines

Bar Chart

Use for: Category comparison

Sales by Region
North America  ████████████████████████████████████████  $50M
Europe         ████████████████████████████              $35M
Asia           ████████████████████                      $25M
Other          ████████████                              $10M
               0    10M   20M   30M   40M   50M

Best practices:

• Horizontal bars for many categories
• Sort by value (not alphabetically)
• Start Y-axis at zero
• Use consistent bar widths

Pie/​Donut Chart

Use for: Parts of whole (max 5 slices)

Market Share
Product A  ████████████████████████████████████████  50%
Product B  ████████████████████                      25%
Product C  ████████████████████                      25%

Best practices:

• Max 5 slices
• Use donut for modern look
• Consider bar chart instead
• Order slices by size

Scatter Plot

Use for: Correlation between two variables

Quantity
100 ┤     ●
     │   ●   ●
 50 ┤ ●   ●   ●
     │ ●   ●   ●
  0 └────────────────────
     0   50  100  150
          Price

Best practices:

• Add trend line
• Color by category
• Identify outliers
• Use transparency for overlapping points

Heatmap

Use for: Two-dimensional data

Sales by Region × Month
         Jan   Feb   Mar   Apr
US      ■■■■■ ■■■■■ ■■■■■ ■■■■■
EU      ■■■■  ■■■■  ■■■■  ■■■■
Asia    ■■■   ■■■   ■■■   ■■■

Best practices:

• Use color scale
• Include legend
• Label both axes
• Consider diverging colors for +/​- data

Histogram

Use for: Distribution of continuous data

Order Value Distribution
$0-10    ████████████████████████████████████████  100
$10-20   ████████████████████████████              80
$20-30   ████████████████████                      60
$30-40   ████████████                              40
$40-50   ██████████                                20
$50+     ██████                                    10

Best practices:

• Choose appropriate bin size
• Show normal distribution if applicable
• Label axes clearly
• Consider density plot for smooth curves

Box Plot

Use for: Distribution with outliers

Salary Distribution
┌─────────────────────────────────────┐
│          ┌───┐                   │
│          │   │                   │
│      ┌───┤   ├───┐               │
│      │   │   │   │               │
│      │   │   │   │               │
│      └───┴───┴───┘               │
└─────────────────────────────────────┘
 Min  Q1  Med  Q3   Max

Best practices:

• Show outliers as points
• Compare multiple box plots
• Label quartiles
• Use for skewed distributions

Color Theory

Color Palettes

Sequential (Ordered Data)

Use for continuous, ordered data.

Viridis: #440154 → #3b528b → #21918c → #5ec962 → #fde725
Blues:   #f7fbff → #deebf7 → #c6dbef → #9ecae1 → #6baed6
Greens:  #f7fcf5 → #e5f5e0 → #c7e9c0 → #a1d99b → #74c476

Diverging (Deviation from Center)

Use for data with meaningful midpoint.

RdYlGn: #a50026 → #d73027 → #f46d43 → #fdae61 → #fee08b
         → #d9ef8b → #a6d96a → #66bd63 → #1a9850 → #006837

Qualitative (Categorical)

Use for distinct categories.

Set1: #e41a1c, #377eb8, #4daf4a, #984ea3, #ff7f00
Set2: #66c2a5, #fc8d62, #8da0cb, #e78ac3, #a6d854

Semantic Colors

Colorblind-Friendly Design

Tips:

• Avoid red-green only
• Use patterns + color
• Test with colorblind simulators
• Use diverging palettes

Safe Palettes:

• Viridis (colorblind-safe)
• ColorBrewer (designed for accessibility)
• Okabe-Ito (8 colorblind-safe colors)

Visual Perception

Preattentive Attributes

Attributes processed instantly (before conscious thought).

Ranking of Visual Attributes

From most to least accurate for quantitative data:

1. Position (most accurate)
2. Length
3. Angle
4. Direction
5. Area
6. Volume
7. Color saturation
8. Color hue (least accurate)

Implication: Use position/​length for precise values, color for categories.

Gestalt Principles

Chart Design Best Practices

1. Direct Labeling

Bad (legend):

[Chart]
Legend: Blue = A, Red = B, Green = C

Good (direct labels):

[Chart with labels on chart]
  Series A ●───●───●
  Series B ●───●───●
  Series C ●───●───●

2. Start Y-Axis at Zero

Bad (truncated):

$100k ┤  ●
 $98k ┤
 $96k ┤  ●
 $94k ┤
       └────────

Good (starts at zero):

$100k ┤  ●
 $50k ┤
   $0 ┤  ●
       └────────

Exception: When zero is not meaningful (e.g., temperature)

3. Sort by Value

Bad (alphabetical):

Zebra    ████████
Apple    ████████████████
Banana   ████████████

Good (by value):

Apple    ████████████████
Banana   ████████████
Zebra    ████████

4. Remove Chart Junk

Bad (cluttered):

[Chart with grid, borders, 3D effects, shadows, gradients]

Good (clean):

[Clean chart with minimal decoration]

5. Use Consistent Scales

Bad (different scales):

Chart A: 0-100
Chart B: 0-1000

Good (consistent scales):

Chart A: 0-100
Chart B: 0-100 (normalized)

Interactive Visualizations

Interactive Features

Tooltip Design

Good tooltip:

┌─────────────────────────────────────┐
│  January 2024                     │
│  ───────────────────────────────  │
│  Revenue: $1,234,567             │
│  Orders: 1,234                   │
│  AOV: $1,000                     │
│  ▲ 12.3% vs previous month       │
└─────────────────────────────────────┘

Zoom and Pan

Before zoom:
┌─────────────────────────────────────┐
│  ●───●───●───●───●───●───●───●  │  ← Full view
└─────────────────────────────────────┘

After zoom:
┌─────────────────────────────────────┐
│           ●───●───●              │  ← Zoomed in
└─────────────────────────────────────┘

Accessibility

WCAG Guidelines

Color Contrast Checker

Good contrast (7:1):

Black text on white background
#000000 on #FFFFFF

Poor contrast (1.5:1):

Light gray on white background
#CCCCCC on #FFFFFF

Color Independence

Bad (color only):

Red = Negative, Green = Positive

Good (color + pattern/​icon):

Red ↓ = Negative, Green ↑ = Positive

Alt Text for Charts

Example:

<figure>
  <img src="revenue-chart.png" alt="Line chart showing revenue
  increasing from $1M in January to $1.5M in June, with a
  12% month-over-month growth rate.">
  <figcaption>Revenue Growth Q1-Q2 2024</​figcaption>
</​figure>

Tools

JavaScript Libraries

Python Libraries

R Libraries

BI Tools

Common Mistakes

1. 3D Charts

Problem: 3D distorts data perception.

Bad:

[3D bar chart with perspective]

Good:

[2D bar chart, flat]

2. Truncated Y-Axis

Problem: Exaggerates differences.

Bad:

$100k ┤  ●
 $98k ┤
 $96k ┤  ●

Good:

$100k ┤  ●
   $0 ┤  ●

3. Too Many Colors

Problem: Confusing, hard to distinguish.

Bad: 10+ colors Good: 3-5 colors max

4. Pie Charts with Many Slices

Problem: Hard to compare.

Bad: 10+ slices Good: Max 5 slices, use bar chart instead

5. Missing Context

Problem: Numbers without meaning.

Bad: "Revenue: $1.2M" Good: "Revenue: $1.2M ▲ 12% vs last month"

6. Rainbow Colors

Problem: No meaning, hard to read.

Bad: Random colors Good: Semantic colors (red=bad, green=good)

7. Small Fonts

Problem: Hard to read.

Bad: 10px font Good: 12px+ font

Mobile-Responsive Charts

Responsive Design

Desktop: Wide charts, side-by-side Mobile: Narrow charts, stacked

Desktop:          Mobile:
┌─────┬─────┐    ┌─────┐
│  A  │  B  │    │  A  │
├─────┼─────┤    ├─────┤
│  C  │  D  │    │  B  │
└─────┴─────┘    ├─────┤
                 │  C  │
                 ├─────┤
                 │  D  │
                 └─────┘

Touch-Friendly

• Minimum tap target: 44×44px
• Large touch areas
• Swipe gestures

Performance

• Optimize images
• Lazy load charts
• Use canvas for many points

Animation and Transitions

Animation Principles

Transition Types

Example: D3.js Animation

// Smooth transition
d3.select('.bar')
  .transition()
  .duration(500)
  .attr('height', newHeight);

Implementation Examples

Chart.js Example

import Chart from 'chart.js/​auto';

const ctx = document.getElementById('myChart');

new Chart(ctx, {
  type: 'line',
  data: {
    labels: ['Jan', 'Feb', 'Mar', 'Apr', 'May'],
    datasets: [{
      label: 'Revenue',
      data: [100, 120, 115, 134, 168],
      borderColor: 'rgb(75, 192, 192)',
      backgroundColor: 'rgba(75, 192, 192, 0.2)',
      tension: 0.1
    }]
  },
  options: {
    responsive: true,
    plugins: {
      legend: {
        display: false
      },
      tooltip: {
        callbacks: {
          label: function(context) {
            return `$${context.parsed.y}K`;
          }
        }
      }
    },
    scales: {
      y: {
        beginAtZero: true,
        ticks: {
          callback: function(value) {
            return `$${value}K`;
          }
        }
      }
    }
  }
});

D3.js Example

import * as d3 from 'd3';

const data = [
  { month: 'Jan', value: 100 },
  { month: 'Feb', value: 120 },
  { month: 'Mar', value: 115 },
  { month: 'Apr', value: 134 },
  { month: 'May', value: 168 }
];

const margin = {top: 20, right: 30, bottom: 40, left: 50};
const width = 800 - margin.left - margin.right;
const height = 400 - margin.top - margin.bottom;

const svg = d3.select('#chart')
  .append('svg')
  .attr('width', width + margin.left + margin.right)
  .attr('height', height + margin.top + margin.bottom)
  .append('g')
  .attr('transform', `translate(${margin.left},${margin.top})`);

const x = d3.scaleBand()
  .domain(data.map(d => d.month))
  .range([0, width])
  .padding(0.2);

const y = d3.scaleLinear()
  .domain([0, d3.max(data, d => d.value)])
  .range([height, 0]);

svg.append('g')
  .attr('transform', `translate(0,${height})`)
  .call(d3.axisBottom(x));

svg.append('g')
  .call(d3.axisLeft(y));

svg.selectAll('.bar')
  .data(data)
  .enter()
  .append('rect')
  .attr('class', 'bar')
  .attr('x', d => x(d.month))
  .attr('width', x.bandwidth())
  .attr('y', height)
  .attr('height', 0)
  .transition()
  .duration(500)
  .attr('y', d => y(d.value))
  .attr('height', d => height - y(d.value))
  .attr('fill', '#4daf4a');

Plotly Example (Python)

import plotly.express as px
import pandas as pd

data = pd.DataFrame({
    'month': ['Jan', 'Feb', 'Mar', 'Apr', 'May'],
    'revenue': [100, 120, 115, 134, 168]
})

fig = px.line(
    data,
    x='month',
    y='revenue',
    title='Revenue Trend',
    labels={'revenue': 'Revenue ($K)', 'month': 'Month'},
    markers=True
)

fig.update_layout(
    yaxis_range=[0, 200],
    hovermode='x unified'
)

fig.update_traces(
    line=dict(width=3, color='#4daf4a'),
    marker=dict(size=8)
)

fig.show()

Seaborn Example (Python)

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

data = pd.DataFrame({
    'region': ['North', 'South', 'East', 'West'],
    'sales': [50, 35, 25, 10]
})

plt.figure(figsize=(10, 6))

ax = sns.barplot(
    data=data,
    x='sales',
    y='region',
    palette='viridis'
)

ax.set_xlabel('Sales ($M)')
ax.set_ylabel('Region')
ax.set_title('Sales by Region')

# Add value labels
for i, v in enumerate(data['sales']):
    ax.text(v + 1, i, f'${v}M', va='center')

plt.tight_layout()
plt.show()

Summary Checklist

Before Creating Visualization

• Understand audience and goal
• Choose appropriate chart type
• Select color palette
• Plan layout
• Consider accessibility

During Creation

• Use direct labeling
• Start Y-axis at zero (if appropriate)
• Sort by value
• Remove chart junk
• Add context

---

## Quick Start

### Basic Chart with Chart.js

```javascript
import { Chart } from 'chart.js'

const ctx = document.getElementById('myChart')
const chart = new Chart(ctx, {
  type: 'line',
  data: {
    labels: ['Jan', 'Feb', 'Mar', 'Apr'],
    datasets: [{
      label: 'Revenue',
      data: [1000, 1200, 1100, 1300],
      borderColor: 'rgb(75, 192, 192)',
      tension: 0.1
    }]
  },
  options: {
    responsive: true,
    scales: {
      y: {
        beginAtZero: true
      }
    }
  }
})

D3.js Basic Bar Chart

import * as d3 from 'd3'

const data = [10, 20, 30, 40, 50]
const svg = d3.select('body').append('svg')
  .attr('width', 400)
  .attr('height', 300)

svg.selectAll('rect')
  .data(data)
  .enter()
  .append('rect')
  .attr('x', (d, i) => i * 50)
  .attr('y', d => 300 - d * 5)
  .attr('width', 40)
  .attr('height', d => d * 5)

---

Production Checklist

• Chart Selection: Choose appropriate chart type for data
• Data Accuracy: Ensure data is accurate and up-to-date
• Accessibility: Charts accessible to screen readers
• Responsive: Charts work on all screen sizes
• Performance: Charts render efficiently with large datasets
• Color: Use color-blind friendly palettes
• Labels: Clear labels and legends
• Context: Provide context and explanations
• Interactivity: Add tooltips and interactions where helpful
• Testing: Test with real data
• Documentation: Document chart purpose and data source
• Updates: Keep charts current with data changes

---

Anti-patterns

❌ Don't: Misleading Y-Axis

// ❌ Bad - Y-axis doesn't start at zero
chart.options.scales.y.min = 90  // Misleading!

// ✅ Good - Y-axis starts at zero
chart.options.scales.y.min = 0  // Accurate representation

❌ Don't: Chart Junk

// ❌ Bad - Too many decorations
chart.options.plugins.legend.display = true
chart.options.plugins.title.display = true
chart.options.plugins.annotation = { /​* decorations */ }
// Too much!

// ✅ Good - Clean, focused
chart.options.plugins.legend.display = true  // Only if needed
// Remove unnecessary decorations

❌ Don't: Wrong Chart Type

// ❌ Bad - Line chart for categories
type: 'line'  // Categories don't have trends

// ✅ Good - Bar chart for categories
type: 'bar'  // Better for comparing categories

---

Integration Points

• Dashboard Design (23-business-analytics/​dashboard-design/) - Dashboard layouts
• KPI Metrics (23-business-analytics/​kpi-metrics/) - Metric visualization
• SQL for Analytics (23-business-analytics/​sql-for-analytics/) - Data queries

---

Further Reading

• Edward Tufte's Books
• D3.js Documentation
• Chart.js Documentation

After Creation

• Test for accessibility
• Verify on mobile
• Check color contrast
• Get feedback
• Iterate based on feedback