1_data_visualization_fundamentals

Sources:

• Excel Data Visualization Course – Guide to Charts & Dashboards
• Data Visualization Course | Data Visualization Tutorial | Data Visualization Explained | Simplilearn

Introduction to Data Visualization

Data visualization is the graphical representation of information and data. Using visual elements like charts, graphs, and maps, data visualization tools provide an accessible way to see and understand trends, outliers, and patterns in data.

Why Data Visualization Matters

• Pattern Recognition: Visual formats make it easier to identify trends and correlations
• Communication: Complex data becomes accessible to broader audiences
• Decision Making: Visual insights support data-driven decision making
• Storytelling: Visualizations help narrate the story behind the data

Types of Data Visualization

Exploratory Data Visualization

• Purpose: Helps understand and discover patterns in data
• Approach: Keep all potentially relevant details visible
• Question: How much detail can we interpret effectively?
• Focus: Discovery and analysis phase

Explanatory Data Visualization

• Purpose: Share findings and tell a story with data
• Approach: Make editorial decisions about emphasis
• Question: What features to highlight vs. eliminate?
• Focus: Communication and presentation phase

Data Types and Field Classifications

Dimensions vs. Measures

Understanding your data fields is fundamental to choosing appropriate visualizations.

Dimensions (Qualitative/Discrete)

• Definition: Descriptive values such as names, dates, categories, and geographical data
• Characteristics:
  • Cannot be aggregated meaningfully
  • Used to categorize and segment data
  • Provide context for measures
• Examples:
  • Product names, customer names
  • Dates, months, years
  • Categories (High/Medium/Low)
  • Geographic locations (City, State, Country)

Measures (Quantitative/Continuous)

• Definition: Numeric values that can be measured and aggregated
• Characteristics:
  • Can be summed, averaged, counted, or calculated
  • Represent magnitude or quantity
  • Support mathematical operations
• Examples:
  • Sales amounts, quantities sold
  • Temperature, weight, height
  • Counts of orders or customers

Discrete vs. Continuous

Discrete Fields (Blue/Categorical):

• Individual, distinct values
• Represented as headers or labels
• Each value is treated separately
• Example: Individual years (2019, 2020, 2021)

Continuous Fields (Green/Numeric):

• Range of values on a scale
• Represented as axes
• Shows progression and flow
• Example: Time as a continuous axis from 2019 to 2021

Chart Types and Their Applications

Column Charts

Best for: Comparing values across categories or showing trends over time

Description: Column charts use vertical rectangular bars of varying lengths to represent different values. Categories are placed on one axis, values on another.

Types:

• Clustered Column: Side-by-side comparison of multiple data series
• Stacked Column: Shows composition within each category
• 100% Stacked Column: Shows proportional contribution of each component

When to Use:

• Comparing values across discrete categories
• Showing data variation over time periods
• Comparing multiple data series simultaneously

Bar Charts

Best for: Comparing discrete values, especially with long category labels

Description: Similar to column charts but use horizontal rectangular bars. The length of each bar corresponds to the magnitude of the value.

Types:

• Clustered Bar: Compares multiple series horizontally
• Stacked Bar: Shows cumulative values
• 100% Stacked Bar: Shows percentage composition

When to Use:

• Category labels are long or numerous
• Comparing rankings
• When horizontal orientation improves readability

Line Charts

Best for: Showing trends and changes over time (time series analysis)

Description: Line charts display information as a series of data points connected by straight line segments.

Types:

• Simple Line: Basic trend visualization
• Stacked Line: Cumulative trends
• Line with Markers: Emphasizes individual data points
• 100% Stacked Line: Shows relative trends

When to Use:

• Tracking changes over continuous time periods
• Comparing trends between multiple variables
• Identifying patterns in sequential data

Area Charts

Best for: Showing magnitude of change over time and cumulative totals

Description: Similar to line charts but with the area below the line filled in. Multiple series can be stacked to show part-to-whole relationships.

Types:

• Simple Area: Single series magnitude
• Stacked Area: Cumulative contribution of multiple series
• 100% Stacked Area: Proportional contribution over time

Benefits:

• Emphasizes volume or magnitude
• Shows cumulative totals visually
• Better than line charts for showing total accumulation

Pie and Donut Charts

Best for: Showing proportional composition of a whole

Description:

• Pie Chart: Circle divided into sectors, each representing a proportion
• Donut Chart: Variation with a hollow center, allowing for multiple series or additional information

When to Use:

• Showing percentage breakdown of a total
• Limited number of categories (ideally < 6)
• Emphasizing proportion rather than precise values

Caution: Difficult for precise comparison; use when approximate proportions suffice

XY Scatter Plots

Best for: Identifying relationships and correlations between two variables

Description: Uses Cartesian coordinates to display values for two variables as points on a plane.

Types:

• Scatter: Individual data points
• Scatter with Smooth Lines: Shows trend curves
• Bubble Chart: Adds third dimension through point size

When to Use:

• Determining correlation between variables
• Identifying clusters or outliers
• Visualizing distribution patterns

Key Insight: The pattern of points reveals the nature of the relationship:

• Positive correlation: Points trend upward left to right
• Negative correlation: Points trend downward
• No correlation: Random scatter

Heat Maps

Best for: Visualizing complex data matrices and identifying patterns

Description: Two-dimensional matrix visualization where values are represented by colors.

Characteristics:

• Uses color intensity to represent values
• Effective for showing relationships between two categorical variables
• Can handle multiple measures simultaneously

When to Use:

• Comparing multiple variables across categories
• Identifying high and low value concentrations
• Visualizing correlation matrices

Tree Maps

Best for: Displaying hierarchical data with proportional representation

Description: Rectangular visualization where each rectangle represents a category, with size proportional to a quantitative value. Rectangles can be nested for hierarchical data.

When to Use:

• Visualizing hierarchical data structures
• Showing part-to-whole relationships with many categories
• Dashboard space optimization

Radar (Spider) Charts

Best for: Comparing multiple quantitative variables across categories

Description: Displays multivariate data as a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point.

When to Use:

• Performance analysis across multiple dimensions
• Comparing entities with multiple attributes
• Quality assessment across various metrics

3.10. Box and Whisker Charts (Box Plots)

Best for: Displaying data distribution and statistical summary

Description: Visual representation of statistical summary including:

• Minimum: Lowest value (excluding outliers)
• 25th Percentile (Q1): First quartile
• Median (50th Percentile): Middle value
• 75th Percentile (Q3): Third quartile
• Maximum: Highest value (excluding outliers)

When to Use:

• Comparing distributions across categories
• Identifying outliers
• Understanding data spread and skewness

3.11. Histograms

Best for: Showing frequency distribution of continuous data

Description: Bar chart variant where bars represent ranges (bins) of data, showing how frequently values occur within each range.

Key Concepts:

• Bins: Intervals that group data values
• Frequency: Count of observations within each bin
• Distribution Shape: Shows if data is normal, skewed, or has other patterns

When to Use:

• Understanding data distribution
• Identifying central tendency and spread
• Detecting outliers or gaps in data

3.12. Pareto Charts

Best for: Identifying the most significant factors (80/20 principle)

Description: Combination of bar and line chart where bars represent individual values in descending order, and the line represents cumulative percentage.

The Pareto Principle: Roughly 80% of effects come from 20% of causes

When to Use:

• Prioritization of issues or factors
• Quality control analysis
• Resource allocation decisions

3.13. Waterfall Charts

Best for: Showing cumulative effect of sequential positive and negative values

Description: Shows how an initial value is affected by a series of intermediate positive or negative values, creating a “waterfall” effect.

When to Use:

• Financial analysis (cash flow, profit/loss breakdowns)
• Inventory tracking
• Step-by-step variance analysis

3.14. Map Charts (Choropleth Maps)

Best for: Geographic data representation

Description: Color-coded maps where regions are shaded based on data values.

When to Use:

• Regional comparisons
• Geographic pattern identification
• Location-based data analysis

Advanced Visualization Techniques

Dual Axis Charts

Purpose: Display multiple data series with different scales on the same visualization

Types:

• Dual Axis with Different Scales: Each axis has independent scale
• Synchronized Dual Axis: Both axes share the same scale
• Combination Charts: Different chart types (e.g., bar + line)

Use Cases:

• Comparing metrics with vastly different magnitudes
• Showing relationships between different measures
• Highlighting correlations

Best Practices:

• Clearly label which axis corresponds to which measure
• Use synchronized axes when comparing similar magnitudes
• Avoid mixing too many different scales

Blended Axis Charts

Difference from Dual Axis:

• Single shared axis for multiple measures
• Both measures use the same scale
• Necessarily have the same chart type for both measures

When to Use:

• When comparing two similar metrics
• When magnitudes are comparable
• For cleaner, less cluttered visualizations

Chart Selection Guidelines

Choosing the Right Chart

Key Selection Criteria

1. Data Type: Categorical vs. continuous vs. hierarchical
2. Number of Variables: Single, dual, or multiple
3. Comparison Type: Part-to-whole, time-based, or correlation
4. Audience Needs: Precision vs. pattern recognition
5. Message Priority: What insight should be most prominent?

Visualization Best Practices

Color Usage

Color Palettes

• Sequential Palettes: For continuous data ranges (light to dark)
• Diverging Palettes: For data with meaningful midpoint
• Categorical Palettes: For distinct, unrelated categories
• Accessibility: Ensure sufficient contrast for colorblind users

Color Guidelines

• Be consistent across related visualizations
• Use color to emphasize, not decorate
• Consider cultural connotations of colors
• Limit palette to 5-7 colors maximum

Clarity and Simplicity

• Remove unnecessary chart elements (redundant labels, excessive gridlines)
• Focus on the key message
• Limit the number of categories displayed
• Use clear, descriptive titles and labels

Accurate Representation

• Avoid distorting data: Use appropriate scales
• Start axes at zero for bar/column charts (when appropriate)
• Consistent intervals: Maintain regular spacing
• Proper proportions: Ensure visual size matches data values

Summary: Data Visualization Workflow

1. Define Objectives: What questions need answering?
2. Understand Data: What types and structures are available?
3. Classify Fields: Identify dimensions vs. measures
4. Select Chart Types: Match visualization to purpose
5. Apply Design Principles: Ensure clarity and consistency
6. Test and Refine: Validate with target audience
7. Deploy and Monitor: Track usage and effectiveness

Key Takeaways

• Choose charts based on purpose, not aesthetics alone
• Understand your data types: Dimensions vs. measures drives visualization choices
• Maintain consistency in design elements
• Prioritize clarity over complexity
• Consider your audience’s needs and expertise level
• Apply the 80/20 rule: Focus on the most impactful visualizations
• Test for accessibility and colorblind-friendly palettes
• Tell a story with your data, don’t just display it