Unit III Foundations of Data Visualization

Data Preprocessing:
Data preprocessing is a crucial step in the data analysis and visualization pipeline.
Raw data collected from various sources often contains:
 Missing values
 Inconsistent formats
 Redundancies
 Noise (irrelevant or erroneous data)
 Unstructured or semi-structured formats
To ensure meaningful, accurate, and efficient visualization or modeling, we prepare
this data by applying a series of cleaning and transformation techniques.

1. Data Cleaning: It is the process of identifying and correcting errors or

inconsistencies in the dataset. It involves handling missing values, removing
duplicates, and correcting incorrect or outlier data to ensure the dataset is accurate
and reliable. Clean data is essential for effective analysis, as it improves the quality
of results and enhances the performance of data models.
  Missing Values: This occur when data is absent from a dataset. You can
either ignore the rows with missing data or fill the gaps manually, with the
attribute mean, or by using the most probable value. This ensures the dataset
remains accurate and complete for analysis.
 Noisy Data: It refers to irrelevant or incorrect data that is difficult for
machines to interpret, often caused by errors in data collection or entry. It
can be handled in several ways:
o Binning Method: The data is sorted into equal segments, and each
segment is smoothed by replacing values with the mean or boundary
values.
o Regression: Data can be smoothed by fitting it to a regression
function, either linear or multiple, to predict values.
o Clustering: This method groups similar data points together, with
outliers either being undetected or falling outside the clusters. These
techniques help remove noise and improve data quality.
 Removing Duplicates: It involves identifying and eliminating repeated data
entries to ensure accuracy and consistency in the dataset. This process
prevents errors and ensures reliable analysis by keeping only unique records.
2. Data Integration: It involves merging data from various sources into a single,
unified dataset. It can be challenging due to differences in data formats, structures,
and meanings. Techniques like record linkage and data fusion help in combining
data efficiently, ensuring consistency and accuracy.
 Record Linkage is the process of identifying and matching records from
different datasets that refer to the same entity, even if they are represented
differently. It helps in combining data from various sources by finding
corresponding records based on common identifiers or attributes.
 Data Fusion involves combining data from multiple sources to create a
more comprehensive and accurate dataset. It integrates information that may
be inconsistent or incomplete from different sources, ensuring a unified and
richer dataset for analysis.
 3. Data Transformation: It involves converting data into a format suitable
for analysis. Common techniques include normalization, which scales data
to a common range; standardization, which adjusts data to have zero mean
and unit variance; and discretization, which converts continuous data into
discrete categories. These techniques help prepare the data for more accurate
analysis.
 Data Normalization: The process of scaling data to a common range to
ensure consistency across variables.
 Discretization: Converting continuous data into discrete categories for
easier analysis.
 Data Aggregation: Combining multiple data points into a summary form,
such as averages or totals, to simplify analysis.
 Concept Hierarchy Generation: Organizing data into a hierarchy of
concepts to provide a higher-level view for better understanding and
analysis.
4. Data Reduction: It reduces the dataset’s size while maintaining key
information. This can be done through feature selection, which chooses the
most relevant features, and feature extraction, which transforms the data into
a lower-dimensional space while preserving important details. It uses
various reduction techniques such as,
 Dimensionality Reduction (e.g., Principal Component Analysis): A
technique that reduces the number of variables in a dataset while retaining
its essential information.
 Numerosity Reduction: Reducing the number of data points by methods
like sampling to simplify the dataset without losing critical patterns.
 Data Compression: Reducing the size of data by encoding it in a more
compact form, making it easier to store and process.
 Advantages of Data Preprocessing:
 Improved Data Quality: Ensures data is clean, consistent, and reliable for
analysis.
 Better Model Performance: Reduces noise and irrelevant data, leading to
more accurate predictions and insights.
 Efficient Data Analysis: Streamlines data for faster and easier processing.
 Enhanced Decision-Making: Provides clear and well-organized data for
better business decisions.

Disadvantages of Data Preprocessing:

 Time-Consuming: Requires significant time and effort to clean, transform,
and organize data.
 Resource-Intensive: Demands computational power and skilled personnel
for complex preprocessing tasks.
 Potential Data Loss: Incorrect handling may result in losing valuable
information.
 Complexity: Handling large datasets or diverse formats can be challenging.

Overview of Data Visualization:

What is Data Visualization?
Data Visualization is the graphical representation of data and information using
visual elements such as:
 Charts, Graphs, Maps, Diagrams, Dashboards
It helps people understand the patterns, trends, and insights in data more quickly
and easily than raw numbers or tables.
Importance/Need of Data Visualization:

1. Data Visualization Simplifies the Complex Data

Large and complex data sets can be challenging to understand. Data visualization
helps break down complex information into simpler, visual formats making it
easier for the audience to grasp. For example in a scenario where sales data is
visualized using a heat map on Tableau states that have suffered a net loss are
colored red. This visual makes it instantly obvious which states are
underperforming.
2. Enhances Data Interpretation
Visualization highlights patterns, trends, and correlations in data that might be
missed in raw data form. This enhanced interpretation helps in making informed
decisions. Consider another Tableau visualization that demonstrates the
relationship between sales and profit. It might show that higher sales do not
necessarily equate to higher profits this trend that could be difficult to find from
raw data alone. This perspective helps businesses adjust strategies to focus on
profitability rather than just sales volume.

3. Data Visualization Saves Time

It is definitely faster to gather some insights from the data using data visualization
rather than just studying a chart. In the screenshot below on Tableau it is very easy
to identify the states that have suffered a net loss rather than a profit. This is
because all the cells with a loss are coloured red using a heat map, so it is obvious
states have suffered a loss. Compare this to a normal table where you would need
to check each cell to see if it has a negative value to determine a loss. Visualizing
Data can save a lot of time in this situation.
4. Improves Communication
Visual representations of data make it easier to share findings with others
especially those who may not have a technical background. This is important in
business where stakeholders need to understand data-driven insights quickly. Let
see the below TreeMap visualization on Tableau showing the number of sales in
each region of the United States with the largest rectangle representing California
due to its high sales volume. This visual context is much easier to grasp rather than
detailed table of numbers.
5. Data Visualization Tells a Data Story
Data visualization is also a medium to tell a data story to the viewers. The
visualization can be used to present the data facts in an easy-to-understand form
while telling a story and leading the viewers to an inevitable conclusion. This data
story should have a good beginning, a basic plot, and an ending that it is leading
towards. For example, if a data analyst has to craft a data visualization for
company executives detailing the profits of various products then the data story
can start with the profits and losses of multiple products and move on to
recommendations on how to tackle the losses.

Best Practices for Visualizing Data:

Effective data visualization is crucial for conveying insights accurately. Follow
these best practices to create compelling and understandable visualizations:
1. Audience-Centric Approach: Tailor visualizations to your audience’s
knowledge level, ensuring clarity and relevance. Consider their familiarity
with data interpretation and adjust the complexity of visual elements
accordingly.
1. Design Clarity and Consistency: Choose appropriate chart types, simplify
visual elements, and maintain a consistent color scheme and legible fonts.
This ensures a clear, cohesive, and easily interpretable visualization.
1. Contextual Communication: Provide context through clear labels, titles,
annotations, and acknowledgments of data sources. This helps viewers
understand the significance of the information presented and builds
transparency and credibility.
1. Engaging and Accessible Design: Design interactive features thoughtfully,
ensuring they enhance comprehension. Additionally, prioritize accessibility
by testing visualizations for responsiveness and accommodating various
audience needs, fostering an inclusive and engaging experience

Tools for Data Visualization:

 Power BI
 Tableau
  Microsoft Excel
 Google Data Studio
 Python (Matplotlib, Seaborn, Plotly)
 R (ggplot2, Shiny)
 D3.js (for web-based interactive graphics)

The Human Brain and Data Visualization:

Data visualization is effective because it matches the way the human brain
naturally understands information. The brain can process images and visuals much
faster than text or numbers. This is why using charts, graphs, and visuals helps
people understand data quickly and easily.
1. Fast Visual Processing
 The brain can understand images and visual patterns almost instantly.
 It takes more time and effort to read numbers or text.
 That's why charts and graphs help people see trends or differences quickly.

2. Pre-attentive Attributes
 These are things our brain notices automatically, even before we think about
them.
 Examples: Color, size, shape, position, and orientation.
 These help highlight important data without needing extra explanation.
Example: A red-colored bar in a bar chart stands out more than the rest.

3. Pattern Recognition
 The human brain is good at finding patterns, trends, and outliers.
  Visualizations like line graphs or scatter plots help users find patterns in data
quickly.

4. Gestalt Principles of Perception

These are simple rules about how our brain groups visual items:

Principle Meaning

Proximity Items close together are seen as related

Similarity Items that look similar are grouped

Continuity Our eyes follow lines or shapes naturally

Closure Our brain fills in missing information to see a complete image

Figure-Ground We focus on the main object (figure) and ignore the background

These principles are used in charts and dashboards to organize information better.

5. Cognitive Load
 The brain has limited capacity to process too much information at once.
 If a visual is too complicated or messy, it becomes hard to understand.
 Simple and clear visuals are easier for the brain to handle.

6. Use of Color and Contrast

 Color helps to draw attention and separate different data points.
 Contrast (e.g., dark vs light) makes important parts stand out.
 Too many colors can confuse the viewer, so use them wisely.
The Shapes of Data:
1. Structured Data
 Data that is organized in rows and columns (like a table).
 Each column has a specific type of data (e.g., name, age, salary).
 Easy to store in databases, Excel, or CSV files.
Example:

ID Name Age Salary

1 Alice 25 50,000

2 Bob 30 60,000

Tools: Excel, SQL, Power BI, Tableau

2. Unstructured Data
 Data that does not follow a clear format.
 It includes text, images, audio, video, or social media posts.
 Harder to store and analyze using traditional tools.
Examples:
 A folder of images
 Customer feedback or product reviews in paragraph form
 YouTube videos or audio recordings
Tools: NLP tools (for text), image processing software, AI tools
3.Semi-Structured Data
 Partially organized data — not as clean as tables, but still has some
structure.
 Uses tags or keys to organize data.
Examples:
 JSON, XML, or log files from websites
4. Time-Series Data (Temporal Data)
 Data that is collected over time (minutes, days, years).
 Helps track trends or changes over time.
Example:

Date Temperature

01-01-
22°C
2024

02-01-
24°C
2024

Visuals Used: Line charts, area charts

5. Spatial or Geospatial Data
 Data that shows location or position on Earth.
 Often includes latitude and longitude or maps.
Example:
 GPS data
 City-wise population data
Visuals Used: Maps, heatmaps

Inputs for Data Visualization:

When creating a data visualization, multiple inputs play a role in shaping how the
final product will look and function. These inputs are critical because they affect
how effectively the data is communicated to the viewer. A well-thought-out input
design can make the difference between a confusing visualization and one that
clearly conveys insights. Let's break down the key inputs into several categories:

1. Data Type
The type of data you're visualizing determines the format, structure, and type of
visualization you should use. There are several common data types:
a. Quantitative Data (Numerical Data):
 Continuous: Data that can take any value within a range, such as
temperature, age, or sales over time. Visualizations for continuous data
typically use line charts, histograms, and scatter plots.
 Discrete: Data that consists of distinct, separate values, like the count of
objects or categories. For discrete data, you might use bar charts or dot
plots.
b. Categorical Data (Qualitative Data):
 Nominal Data: Data that consists of categories without any inherent order.
Examples include colors, names, or types of fruit. Visualizations such as pie
charts, bar charts, or word clouds are common for nominal data.
 Ordinal Data: Data with a clear, ranked order but no fixed intervals.
Examples include survey ratings (e.g., "Poor," "Fair," "Good").
Visualizations for ordinal data might use ranked bar charts or stacked bar
charts.
c. Temporal Data (Time-based Data):
 This data involves time as a key variable (e.g., timestamps, dates). Time-
based visualizations require special handling of date and time series, such as
time series plots, line charts, or area charts.
d. Geospatial Data (Location-based Data):
 Data that represents geographic locations, such as coordinates (latitude,
longitude). You would typically use maps, choropleth maps, or bubble
maps for geospatial data to show patterns by location.
e. Multidimensional Data:
 Data with more than two variables. These can include combinations of
quantitative and categorical data. Tools for visualizing multidimensional
data include scatter plots with multiple variables, heatmaps, parallel
coordinate plots, or radar charts.

2. Design Inputs
Design plays a significant role in making data visualization intuitive and easy to
understand. Several design elements need to be carefully considered:
a. Layout and Structure:
 The arrangement of elements in a visualization must be logical and easy to
follow. A poor layout can confuse the viewer.
 For example, if you're visualizing sales over time, having the time axis on
the horizontal axis is standard practice. Make sure there’s enough spacing
between elements, and keep visualizations uncluttered.
 Hierarchy: Decide what information should be most prominent, and ensure
this information stands out with appropriate size, color, or placement.
 Gridlines & Axes: Choose the appropriate number and style of gridlines for
context without overcrowding the visualization.
b. Aesthetics (Colors, Shapes, and Fonts):
 Colors: Colors can be used to differentiate between categories, highlight
trends, or represent values. But color should be chosen carefully to avoid
confusion. Consider colorblind users by choosing color schemes that are
universally distinguishable.
o Use color to represent categories or values.
o Avoid overuse of color; maintain clarity and simplicity.
 Shapes: The choice of shape (e.g., bars, dots, lines, areas) influences how
data is perceived. Use simple and consistent shapes to represent data points.
  Fonts: Font choices should be clear and legible. Titles and axis labels need
to stand out, while data annotations should be subtle. Use fonts consistently
for readability.
c. Interactive Elements:
 For dashboards or interactive visualizations, it’s important to think about
elements such as tooltips, filters, and dropdowns that allow users to
explore the data in different ways.
 Interaction adds flexibility, but excessive interactivity can overwhelm users.
Make sure interactive components are intuitive.

3. Contextual Inputs
Context is crucial because it guides the audience's understanding of the data and
ensures the visualization meets its purpose. Several contextual inputs must be
considered:
a. Audience:
 Who is viewing the data? Understanding the target audience's knowledge,
expertise, and expectations helps in deciding how to present the data.
o Executives may prefer high-level summaries and clear KPIs, while
data analysts may want more detailed charts with complex insights.
 Audience’s Context: What do they already know? Do they need
background information to understand the data, or can you assume some
level of familiarity with the subject matter?
b. Purpose and Message:
 What is the main goal of the visualization? Are you trying to show a trend
over time, highlight a comparison, or demonstrate the relationship
between variables?
 Tell a Story: The visualization should guide the viewer toward
understanding a clear message. Avoid presenting too much data in one chart,
which can overwhelm or confuse the audience.
c. Data Sources and Metadata:
 Providing context about where the data comes from, when it was collected,
and any potential biases is important. This can be presented in the source
metadata, or a footnote in the visualization.
 Attribution: Citing sources adds credibility to the visualization and allows
the audience to understand the reliability of the data.

4. Technical Inputs
These inputs are related to the tools and technologies used to create the
visualization and how data is handled and processed:
a. Data Format and Structure:
 How is the data stored? (e.g., CSV, Excel, SQL databases, APIs)
 Understanding the structure of the data helps in choosing the right tool for
visualization and cleaning/preprocessing. For example, structured data can
be directly loaded into tools like Tableau or PowerBI, while unstructured
data might require more preparation in Python or R before it’s ready for
visualization.
b. Software and Tools:
 The choice of tools for creating data visualizations plays a critical role.
Popular tools include:
o Tableau and PowerBI for business analytics and dashboard creation.
o D3.js for custom, interactive web visualizations.
o Matplotlib, Seaborn, and Plotly for creating visualizations in
Python.
o ggplot2 in R for statistical and data-driven visualizations.
 The tool should align with the complexity of the data and the intended
audience. For example, PowerBI and Tableau are great for non-technical
users, while D3.js provides full flexibility for advanced users.
c. File Size and Performance:
 Consider the file size and performance issues, especially for large datasets.
A large dataset might require optimization techniques like sampling or
aggregation to avoid performance lags in visualization tools.
 Interactive visualizations may need to be optimized for web performance,
especially if you are working with web-based tools.

5. Time and Budget Inputs

a. Time Constraints:
 Visualizations for time-sensitive reporting might need to be created quickly,
with a focus on clarity over complexity.
 In contrast, long-term projects can afford to include more in-depth analyses
and visualizations that provide richer insights.
b. Budget:
 Some visualization tools and software require licensing fees (e.g., Tableau,
PowerBI). If the budget is limited, you may need to choose open-source
options like Google Data Studio or Matplotlib.
 The budget also influences whether you can invest in hiring a designer for a
high-end aesthetic or rely on built-in templates within software.
 Types of Visualizations: Cognitive vs Perceptual
In the world of data visualization, the goal is to present data in a way that’s
easy to understand and helps people make decisions. When designing
visualizations, two important aspects to consider are cognitive and
perceptual factors. These two terms relate to how we process and interpret
information through visuals, and understanding the difference between them
can help you design more effective visualizations.

1. Cognitive Design in Data Visualization

Cognitive design focuses on how the human brain processes and
understands information over time. It’s all about how we think,
remember, and reason when we look at data.
When we talk about cognitive in visualizations, we mean how the
mental effort required to interpret a visualization can influence its
effectiveness. Cognitive design considers things like:
 Clarity: How easy is it to understand what the data is saying?
 Memory: Does the visualization help us remember key points or
trends?
 Logical Thinking: Does it allow us to make inferences and draw
conclusions easily?
The goal of cognitive design is to reduce the mental load on the viewer,
allowing them to quickly grasp the data's meaning without getting
overwhelmed.
Examples of Cognitive Design in Visualizations:
1. Bar Charts for Comparison:
o When you want to compare quantities across categories (e.g.,
sales of different products), bar charts are easy to understand.
They make it simple for our brains to compare the heights of
bars and make quick judgments.
o Why it's cognitive: Bar charts reduce cognitive effort by
organizing the data in a way that’s easy to compare visually.
2. Line Graphs for Trends:
o Line graphs help us understand how something changes over
time. For example, if you’re showing the stock prices over a
month, the viewer can easily track the upward or downward
movement.
o Why it's cognitive: The line graph leverages our ability to
follow changes over time, making it easier to identify
patterns and trends.
Cognitive Principles for Effective Visualization:
 Simplify Complex Data: Don’t overload the viewer with too much
information at once. Use clear, simple visualizations that highlight
key points.
 Use Familiar Formats: People are used to certain formats like bar
charts or line graphs. Stick to what people already know to reduce
cognitive load.
 2. Perceptual Design in Data Visualization
Perceptual design is all about how our brains physically perceive
visual elements. This is more about how we see things and
interpret visual features like color, size, shape, and position.
Perceptual design aims to optimize visual elements so that they are
easily and quickly understood.
When we talk about perceptual in visualizations, we’re concerned
with how human vision works. This includes:
 Color: How colors stand out or convey meaning.
 Size and Shape: How different sizes or shapes of elements
communicate different values or categories.
 Position: How the positioning of elements on the graph helps us
interpret relationships (e.g., higher points might indicate greater
values).
 Visual Encoding: This refers to how we represent data visually
using things like color, shape, size, etc.
The goal of perceptual design is to take advantage of the way our
eyes and brain work together to interpret visual elements. By doing
this, we can make data easier to perceive and understand at a
glance.
Examples of Perceptual Design in Visualizations:
1. Color to Represent Categories:
o Different colors are often used to represent different
categories or values. For example, in a pie chart showing
market share, each segment could be assigned a different
color.
 o Why it’s perceptual: Colors are pre-attentive features,
meaning we notice them instantly without having to think
about it. Our brains can distinguish between different colors
quickly, making them an excellent tool for showing different
groups or categories.
2. Size to Represent Quantity:
o In bubble charts, the size of each bubble can represent the
magnitude of a variable (e.g., sales volume or population
size). Larger bubbles indicate larger quantities.
o Why it’s perceptual: We are good at perceiving differences
in size quickly and easily. Larger items seem more important,
so size can effectively convey quantitative differences.
3. Position to Represent Data:
o In scatter plots, the position of a point on the graph (its X
and Y coordinates) can represent two different variables,
making it easy to see relationships and correlations.
o Why it’s perceptual: The brain naturally understands spatial
relationships—how far one point is from another. So, the
position of data points on the X and Y axes is immediately
meaningful to us.
Perceptual Principles for Effective Visualization:
 Use color wisely: Choose colors that are easy to differentiate and
don’t confuse the viewer. Avoid using too many similar colors,
which can make the visualization confusing.
 Leverage size and position: People can easily spot differences in
size and the position of objects. Use this to your advantage to
highlight key data points.
Aspect Cognitive Design Perceptual Design

How we physically
How we think, reason, and
Focus perceive visual elements
process information.
(like color, size, position).

Reduce mental load, make Optimize how we notice

Goal data easy to understand and interpret visual
and reason about. elements quickly.

How quickly the viewer

How easily the viewer can
Primary can detect visual
understand and process
Concern differences (color, size,
data.
shape, etc.).

Using bar charts to Using color to differentiate

Example compare values across between categories or
categories. values.

Focuses on reducing the Focuses on leveraging how

Cognitive Load mental effort needed to our brains naturally
interpret the data. interpret visual features.

Clarity, simplicity, and Color, size, position, shape,

Key Elements
logical structure. and visual encoding.

Aimed at clear Aimed at quick recognition

Effectiveness communication and easy and easy differentiation of
decision-making. data points.

Typical Use Showing trends, Highlighting categories,

Cases comparisons, and quantities, or data patterns
 relationships in data. at a glance.

Uses visual features like

Perceptual Not directly relevant to
color, size, shape, position,
Features Used cognitive design.
and texture.

Example in Line charts for trends, bar Color-coded pie charts,

Visualization charts for comparison. bubble sizes for quantities.

Involves reasoning and Involves quick recognition

Mental Process deeper thought about the of visual patterns and
data. differences.

5 Big Data Visualization Categories:

Temporal: Data related to time, represented through line charts,
timelines, etc.
Hierarchical: Data with a parent-child structure, displayed in tree
maps, dendrograms, or sunburst charts.
Network: Data that represents relationships between entities, often
displayed through network graphs.
Multidimensional: Data involving more than two variables, often
visualized in scatter plots with more than two dimensions (e.g., 3D
plots).
Geospatial: Data with geographic information, visualized on maps,
including choropleth maps or geospatial heatmaps.
1. Temporal Visualization:
Temporal visualization is a method of representing data that changes
over time. It's essential for analyzing time-dependent data, allowing us
to observe trends, fluctuations, and patterns across different time
intervals (e.g., seconds, hours, days, months, or years). Time, being a
continuous variable, is typically represented on the x-axis, while the data
of interest (such as sales, temperature, or stock prices) is plotted on the
y-axis.
Temporal visualization is widely used in various fields such as business,
finance, healthcare, and scientific research to help make sense of data
trends and patterns that evolve over time.
Common Types of Temporal Visualizations:
Line chart, area, bar chart , Heatmaps, Gantt Charts, Time Series Plots.

Line Charts
 Description: A line chart is one of the most common forms of
temporal visualization. It uses a line to connect individual data
points plotted over time.
 Use Case: Line charts are ideal for showing trends and changes
over a continuous period.
 Example: Tracking the daily temperature in a city over a month.
The x-axis would represent the days of the month, and the y-axis
would show the temperature.
Area Charts
 Description: An area chart is similar to a line chart but with the
area beneath the line filled in. This emphasizes the volume of data
and is particularly useful for showing the cumulative impact over
time.
 Use Case: Area charts are useful when you want to visualize the
magnitude of change over time, especially when comparing
multiple categories or datasets.
 Example: Showing the cumulative sales of different products over
time. The area chart would allow you to visualize both individual
and cumulative sales growth.
Why Temporal Visualizations Are Important
 Trend Identification: Temporal visualizations help you spot increasing or
decreasing trends over time, allowing you to make predictions or decisions
based on those patterns.
 Seasonality and Cycles: Many types of data exhibit seasonal or cyclical
behavior, such as sales spikes during the holiday season or weather patterns
throughout the year. Temporal visualization makes these cycles easy to
identify.
 Anomaly Detection: Temporal visualizations allow you to quickly spot any
irregular spikes or drops in the data, which can indicate issues such as
product failures, system crashes, or market disruptions.
 Forecasting: Temporal data is key to predictive analysis. By observing past
trends, businesses and researchers can forecast future behavior (e.g., sales,
stock prices, or demand for resources).

2. Hierarchical Visualization:
Hierarchical visualization is a method used to represent data with a tree-like
structure, where elements have parent-child relationships. This type of
visualization is ideal for showing how items are organized across multiple levels
— from the top (root) down to more specific sub-levels (leaves).
It's commonly used in data structures, organizational charts, file systems,
biological taxonomy, and any dataset where information can be categorized into
nested groups.
Key Concepts of Hierarchical Data
 Hierarchy: A system where elements are ranked above or below one
another.
 Parent Node: A higher-level category or group.
 Child Node: A more specific item or subgroup within a parent.
 Root: The topmost element in the hierarchy.
 Leaf: The lowest level item with no children.
Common Types of Hierarchical Visualizations:
Tree Diagrams, Treemaps, Dendrograms, Sunburst Charts

Tree Diagrams
 Description: Displays data as a branching structure from the root node to
the leaf nodes.
 Use Case: Visualizing organizational charts, decision trees, and
classification systems.
 Example: A company org chart with the CEO at the top, branching out to
department heads, then team leaders.
Treemaps:
 Structure: Nested rectangles representing hierarchy. Size and color show
additional metrics.
 Best for: Space-efficient visualization of large hierarchies.
 Example:
o Visualizing disk usage: Folder size → subfolder size → file size.
o Daily Food Sales..
Sunburst Chart
 Structure: Circular form of a treemap with levels in concentric circles.
 Best for: Visualizing hierarchy where sequence or levels are important.
 Example:
3. Network Visualization:
Network visualization is a way to display relationships (connections) between
entities (nodes). It shows how things are linked, influence each other, or interact in
a system.
Think of it like a web or a map of connections—where the nodes are entities
(people, devices, concepts), and the edges are the connections (friendships, data
flow, communication).
Key Elements
 Nodes (Vertices): Represent entities (e.g., people, computers, websites).
 Edges (Links): Represent relationships or interactions between nodes (e.g.,
messages sent, hyperlinks).
 Directed vs. Undirected:
o Directed edges show one-way relationships (e.g., A → B).
o Undirected edges show two-way connections (e.g., A — B).
 Weighted Edges: Some edges have a strength or frequency (e.g., how often
two people communicate).
Example:
A Node-Link Diagram is a type of network visualization that shows entities
(nodes) and the connections (links or edges) between them. It’s one of the most
intuitive and widely used ways to represent relationships or interactions in data.
Think of it like a map where points (nodes) are connected by lines (links), often
used to explore who is connected to whom, or what is connected to what.
 4. Multidimensional Visualization:
Multidimensional visualization is a technique used to represent datasets that have
more than three variables (dimensions). Since we can't visualize beyond 3D
naturally, these methods encode extra dimensions using color, size, shape, position,
or interactivity to help users analyze complex, high-dimensional data.
In simple terms: It’s a way to "see" lots of variables at once in a meaningful visual
form.
Why Use Multidimensional Visualization?
 To explore relationships, correlations, clusters, or outliers in complex
datasets.
 To compare multiple attributes simultaneously (e.g., sales, profit, region,
customer type).
 To reduce dimensionality (e.g., using PCA or t-SNE) and make hidden
patterns visible.
Popular Types of Multidimensional Visualizations:
1.Parallel Coordinates Plot
 Structure: Vertical axes for each variable; lines connect a record's value
across axes.
 Best for: Spotting patterns, clusters, and outliers.
 Example:

2.Radar Chart (Spider Plot)

 Structure: Each axis radiates from a center point, one per variable.
 Best for: Comparing multiple features across entities.
 Example:
o Comparing three smartphones on screen size, battery, speed, and cost.
 5. Geospatial Visualization:

Geospatial visualization is the process of mapping and analyzing data that has a
geographic or spatial component. It uses maps, charts, and symbols to represent
where things happen, allowing patterns tied to location or geography to emerge
clearly.

In simple terms, it's about visualizing data on a map—like plotting store

locations, population density, or crime hotspots across a city.

Why Use Geospatial Visualization?

 To detect spatial patterns, trends, and clusters

 To understand how location affects outcomes
 To support decisions in urban planning, logistics, marketing, disaster
response, etc.

Common Types of Geospatial Visualizations:

1. Choropleth Map

 Description: Areas (like countries or districts) are shaded based on a

value.
 Use Case: Showing population density, unemployment rates, election
results.
 Example: map shaded by COVID-19 cases per state.
2. Heatmap (Spatial Heatmap)

 Description: Uses color gradients to show concentration or intensity

over an area.
 Use Case: Visualize foot traffic, crime intensity, or wildfire hotspots.
 Example: Heatmap of crime in different cities.

Practicing Good Ethics in Data Visualization:

What Are Data Visualization Ethics?

Ethical data visualization means presenting data truthfully, clearly, and

responsibly—without misleading or manipulating the audience. It's not just
about making charts look good—it's about ensuring accuracy, fairness, and
transparency in how data is shown.

Ethics in data visualization = telling the truth with visuals, not just telling a
compelling story.

Key Points for Ethical Data Visualization:

1. Be Honest and Accurate

  Don’t lie with numbers: Make sure your charts don’t trick people into
thinking something is more important or different than it actually is.
 Example: If you’re making a bar chart, make sure the Y-axis starts at 0 so
the bars don’t look bigger than they really are.

2. Give Enough Information

 Provide context: Always tell the viewer where the data comes from, what
time period it covers, and any other important details.
 Example: If you’re showing data about COVID-19, say if it’s the total
number of cases in one month or the total number of cases in a year.

3. Be Transparent

 Show where your data came from: Always say where you got your data
and whether there are any potential issues (like missing data).
 Example: If you use data from a survey, mention how many people
answered the survey and whether it’s just a small group or a bigger sample.

4. Don’t Select Only the Good Data

 Don’t leave out important information just to make a point look better.
 Example: If you’re comparing two years of sales, don’t just show the good
years; include all the years so people can see the whole picture.

5. Don’t Use Tricky Design

 Keep it simple: Avoid using designs that make things look misleading (like
3D charts or weird colors that confuse the numbers).
 Example: Don’t use a 3D bar chart that makes the bars look bigger than
they actually are.

6. Respect Privacy

 Don’t use personal information in your visualizations unless you have

permission and make sure it’s anonymized when necessary.
 Example: Don’t show the names or exact details of individuals if you’re
working with personal data.
Ineffective Visuals and How to Improve Them:
What Are Ineffective Visuals?

Ineffective visuals are charts, graphs, or maps that fail to communicate the
data clearly or mislead the viewer. They can confuse, distract, or give a false
impression about the data.

Common Problems with Ineffective Visuals

1. Misleading Scales

 Problem: If the axis on a chart doesn’t start at 0, it can make small

differences look huge, misleading people about the significance.
 Example: A bar chart shows a difference in sales between two
products. If the Y-axis starts at 90 instead of 0, the bar for Product A
will look much taller than Product B, even if the actual difference is
small.
 How to Improve: Always start the axis at 0 to show the true
proportions.

2. Overcrowded Charts

 Problem: When too much information is crammed into one chart, it

can be overwhelming and hard to understand.
 Example: A pie chart with 20 slices becomes difficult to read and
doesn’t make the data clear.
 How to Improve: Keep it simple—limit the number of categories, or
break complex data into smaller charts.

3. Using 3D Effects

 Problem: 3D charts or graphs can distort the data, making it hard to

accurately compare values.
 Example: A 3D bar chart might make the bars look bigger or smaller
than they are because of the perspective.
 How to Improve: Stick to 2D charts for simplicity and accuracy.

4. Inconsistent Colors
  Problem: Using too many colors or confusing color schemes can
make the chart hard to read or understand.
 Example: A bar chart uses bright red for some categories, and dull
gray for others, making it difficult to compare.
 How to Improve: Use consistent, simple colors. Use color sparingly
and make sure it's clear what each color represents (e.g., blue for one
category, orange for another).

5. Too Much Text or Labels

 Problem: Including too many labels or text in a chart can clutter the
visual and make it harder to focus on the main message.
 Example: A line graph with a lot of numbers on the grid and too
many data points labeled can confuse the viewer.
 How to Improve: Simplify the text, and only show the most
important data or labels. Use tooltips or hover features for extra
details.

6. Wrong Chart Type

 Problem: Choosing the wrong chart type for your data can confuse
the message you’re trying to communicate.
 Example: Using a pie chart to show changes over time, which is
better suited for a line chart.
 How to Improve: Choose the chart that best suits your data—e.g.,
line charts for trends over time, bar charts for comparisons.

How to Make Visuals More Effective:

1. Use a Clear, Proper Scale

 Always start the Y-axis (or any axis) at zero to avoid misleading the
viewer. This shows true proportions and comparisons.

2. Simplify Your Design

 Don't overcrowd charts with too many categories or too much data.
Limit the amount of information to what’s necessary to tell your
story.

3. Avoid 3D Effects
  Stick with 2D charts. They are easier to read and don’t distort data,
especially when comparing values.

4. Use Simple, Consistent Colors

 Choose colors that are easy to distinguish. Limit the number of colors
and be consistent in your color scheme across charts.

5. Minimize Text

 Reduce labels and focus on the key takeaways. If necessary, use

tooltips or legends to provide additional information.

6. Pick the Right Chart

 Choose the best chart type for your data.

o Bar charts for comparisons,
o Line charts for trends over time,
o Pie charts for parts of a whole (but limit categories to 5-6),
o Heatmaps for intensity patterns.

principles of visual perception:

The principles of visual perception refer to how humans interpret and
process visual information. When it comes to data visualization,
understanding these principles is crucial because it helps designers create
visuals that are easier to understand and that communicate the intended
message effectively.

Principles of Visual Perception in Data Visualization

In data visualization, understanding how humans perceive visual information

is key to designing effective and easily interpretable visuals. The principles
of visual perception help ensure that the data is presented in a way that the
viewer can quickly and accurately understand. Here’s a detailed explanation
of each principle:
1. Figure-Ground Relationship

 Definition: This principle refers to how the human brain distinguishes

between the object of focus (the figure) and the background (the
ground). In a data visualization, the "figure" is the key data or visual
element, and the "ground" is the background or neutral space.
 Application: The main data points (figures) need to stand out clearly
against the background (ground). If the background and data are too
similar in color or contrast, the important elements can become hard
to distinguish.

Example: In a bar chart, the bars (figure) should stand out against a
contrasting light-colored background (ground) to draw attention to the data.

2. Gestalt Principles

Gestalt psychology explains how people organize visual information based

on certain principles. These are critical when structuring data visualizations
to make them intuitive and easy to understand.

 Proximity: Elements that are close together are perceived as related.

o Application: Grouping related data points or categories
together in a visualization (like clustered bars in a bar chart)
helps the viewer perceive them as part of the same group.
o Example: In a pie chart, slices that represent similar categories
should be placed closer together.
 Similarity: Visual elements that are similar (in color, shape, size) are
perceived as related.
o Application: Use consistent color or shapes for data points that
share common attributes.
o Example: Bars representing the same category in a bar chart
should have the same color.
 Closure: People tend to fill in gaps or perceive incomplete shapes as
complete.
o Application: Incomplete or segmented lines (like in line charts)
are still perceived as a continuous trend, helping to convey the
idea of a trend even when data is missing.
o Example: A broken line on a graph still feels continuous due to
closure.
  Continuity: People prefer to see smooth, continuous patterns rather
than abrupt changes or interruptions.
o Application: Design charts and graphs with continuous lines or
smooth transitions to maintain flow and coherence.
o Example: A smooth, continuous line graph is easier to follow
than a jagged, erratic one.

3. Color Perception

 Definition: Color plays a significant role in how we perceive data and

interpret information. It can be used to highlight, group, or distinguish
data points.
 Pre-attentive Attributes: The brain processes certain visual
properties like color, size, and shape very quickly and automatically
before conscious attention.

Application: Use contrasting or attention-grabbing colors for important data

points. Color also helps convey meaning (e.g., red for negative trends, green
for positive).

Example: In a bar chart, using a bright color for high values and a dull color
for low values helps the viewer immediately identify trends.

4. Size and Shape Perception

 Definition: The size and shape of visual elements influence how we

perceive them. Larger elements tend to attract more attention, and
shapes can be used to signify different categories or types of data.

Application: Larger bars, points, or areas are often perceived as more

significant, so make sure to emphasize important data points with larger
sizes.

Example: In a bubble chart, the size of the bubble represents the magnitude
of a particular variable. Larger bubbles are interpreted as having greater
significance.
5. Depth Perception

 Definition: Humans can perceive depth and understand the spatial

relationship of objects in a visual space. This helps us understand how
objects relate to each other in three dimensions.

Application: Depth can be used in visualizations (e.g., 3D graphs) to

represent complex data. However, if not done carefully, it can distort or
make it difficult to compare values accurately.

Example: A 3D pie chart can create a sense of depth, but the perspective
might make it harder to compare slice sizes accurately, so 2D representations
are often preferred for clarity.

6. Attention and Focus

 Definition: The human brain can only focus on a limited amount of

information at a time. Effective data visualizations should guide the
viewer's focus to the most important data without overwhelming them.

Application: Use design techniques like contrast, highlighting, or spacing to

direct attention to the key insights and prevent visual overload.

Example: In a dashboard, key performance indicators (KPIs) are often

highlighted with bold colors or larger font sizes to immediately draw
attention to them.

Color as a Pre-Attentive Attribute in Data Visualization

Pre-attentive attributes are visual properties that our brains process
automatically and very quickly, even before conscious attention is applied.
These attributes help us rapidly detect and interpret visual information.
Color is one of the most powerful pre-attentive attributes in data
visualization.

What is Color as a Pre-Attentive Attribute?

Color is considered a pre-attentive attribute because it can immediately
grab our attention and allow us to process visual information with minimal
cognitive effort. The brain processes color quickly, allowing us to identify
key patterns, trends, and relationships in data without having to focus
intently.

How Color Works as a Pre-Attentive Attribute:

1. Fast Detection:
o The brain can detect differences in color almost instantaneously.
For example, if two bars in a bar chart are different colors, our
eyes will quickly notice the contrast, even if we haven't
explicitly looked for it.
o This ability makes color very useful for drawing attention to
key data points or trends without requiring detailed analysis.
2. Color for Categorization:
o Color is often used to group or categorize data points. By
using different colors for different categories or groups, viewers
can immediately distinguish between them.
o Example: In a pie chart, each slice representing a different
category can be colored differently. This allows viewers to
easily identify and compare categories at a glance.
3. Emotional and Psychological Impact:
o Color not only helps with distinguishing data, but it can also
influence the emotional tone of a visualization. For example:
 Red can signal urgency, warning, or negativity.
 Green often represents positive outcomes or growth.
 Blue can convey calm, stability, or neutrality.
o Understanding the psychological impact of color can help
reinforce the message the data visualization is trying to
communicate.
4. Highlighting Key Data:
o By using contrasting colors, important data points or trends
can be highlighted, guiding the viewer's attention to the most
relevant information.
o Example: In a line chart, a significant data point might be
highlighted in a bright color (e.g., red), making it easy to spot,
while the rest of the data remains in a neutral color (e.g., grey).

Practical Applications of Color as a Pre-Attentive Attribute:

 1. Differentiating Data:
o Example: In a bar chart showing sales data across regions, each
region can be assigned a unique color. This instantly tells the
viewer which region corresponds to which bar.
2. Indicating Trends or Comparisons:
o Example: In a line chart representing stock prices over time,
different colors can be used to represent different companies.
This allows the viewer to quickly compare trends for each
company.
3. Creating Hierarchy:
o Example: In a dashboard, color can be used to signify
importance or urgency, with higher priority metrics (e.g., sales
figures, KPIs) displayed in bold or vibrant colors, while less
important metrics are shown in more neutral tones.
4. Signaling Relationships and Divergences:
o Example: In a heatmap, different colors might be used to
represent different ranges of values (e.g., dark red for high
values, dark blue for low values). This makes it easy for the
viewer to quickly understand patterns or outliers in the data.

Best Practices for Using Color in Data Visualization:

1. Limit the Number of Colors:

o Too many colors can confuse the viewer. Stick to a limited
color palette to avoid overwhelming the audience and to ensure
clarity.
2. Use Contrasting Colors for Clarity:
o Ensure there's enough contrast between colors used for different
elements. If the colors are too similar, it may be difficult for the
viewer to differentiate between them.
3. Maintain Consistency:
o Use colors consistently throughout the visualization. For
example, if blue represents one category in one part of a chart,
it should represent the same category elsewhere in the same
chart.
4. Consider Colorblindness:
o Approximately 8% of men and 0.5% of women have some form
of color blindness. Use color combinations that are
distinguishable for colorblind users (e.g., avoid using red and
 green together). Tools like ColorBrewer can help you choose
color palettes that are accessible to everyone.
5. Use Color for Meaning:
o Color should have a clear meaning in the context of the data.
For example, a red color can represent negative values (such as
losses), while green might represent positive values (such as
gains).

Example of Color as a Pre-Attentive Attribute:

Imagine a line chart showing the sales growth of different products over
time. Each product has a line, and the lines are each assigned a distinct
color:

 The blue line represents Product A.

 The green line represents Product B.
 The red line represents Product C.

Here, the use of color allows the viewer to immediately differentiate

between the products, without needing to focus on the legend or axis labels.
They can easily spot which product is performing the best at a given time
based on the color of the line.

Strategic Use of Contrast in Data Visualization:

Contrast is a powerful design element in data visualization that helps to
make certain data points or trends stand out and guides the viewer’s
attention. In the context of data visualization, contrast refers to the
difference in visual properties such as color, size, shape, brightness, and
texture between elements in a graphic.

Why Contrast is Important in Data Visualization:

1. Focus: Contrast draws attention to specific parts of the data that are
most important or require the viewer’s focus.
2. Clarity: Proper contrast helps make the data easier to read and
interpret, avoiding confusion between different data elements.
3. Hierarchy: Contrast can help establish a visual hierarchy, directing
the viewer’s eye to the most important information first.
 4. Accessibility: Using contrast effectively makes a visualization more
accessible, especially for individuals with visual impairments (such as
color blindness).

Types of Contrast Used in Data Visualization:

1. Color Contrast:
o Definition: The difference between two or more colors. High
color contrast makes elements stand out against each other.
o Application: Use contrasting colors to highlight key data
points, trends, or categories.
o Example: In a bar chart, the bars representing important data
could be a bright color (e.g., red) while less important bars use
a muted color (e.g., grey). This allows the key data to stand out
immediately.
2. Size Contrast:
o Definition: The difference in the size of visual elements, such
as bars, lines, or dots.
o Application: Larger elements naturally grab more attention.
Use size contrast to emphasize important data points.
o Example: In a scatter plot, larger circles can represent more
significant values, such as higher sales or larger quantities,
while smaller circles represent less important data.
3. Shape Contrast:
o Definition: The difference in shapes used for visual elements.
o Application: Different shapes can be used to distinguish
between data series or categories in a chart.
o Example: In a line chart, you could use a solid line for one
category and a dashed line for another to create contrast
between the two.
4. Brightness Contrast:
o Definition: The difference in lightness and darkness of visual
elements.
o Application: Brighter elements naturally attract attention, so
use bright elements for important data and darker elements for
less important details.
o Example: In a heatmap, high values might be represented by
bright red, while low values are shown in dark blue or black,
creating contrast that allows for easy pattern recognition.
5. Text Contrast:
 o Definition: The contrast between text and background.
o Application: Ensure text stands out from the background by
using high contrast (e.g., black text on a white background).
This makes labels, titles, and annotations easier to read.
o Example: The title of a chart could be in a bold, large font with
high contrast to the background to make it clear and readable at
a glance.

Tools for Visualizing: Power BI and Tableau:

Both Power BI and Tableau are two of the most popular tools used for data
visualization and business intelligence. They allow users to connect to
various data sources, create interactive visualizations, and generate insights
from complex datasets. Here's a breakdown of both tools:

1. Power BI

Overview:

Power BI is a powerful business analytics tool developed by Microsoft. It

provides interactive data visualizations and business intelligence capabilities
with an easy-to-use interface. It integrates seamlessly with various Microsoft
products (Excel, Azure, etc.) and is designed to work with cloud-based and
on-premises data sources.

Key Features:

1. User-Friendly Interface:
o Power BI is known for its simple and intuitive drag-and-drop
interface, which makes it accessible to both beginners and
experienced users.
2. Data Integration:
o Power BI can integrate with various data sources like Excel,
SQL Server, Google Analytics, Salesforce, and more. It also
supports cloud-based data connections, including Azure and
Power BI Service.
3. Interactive Dashboards:
 o You can create interactive reports and dashboards that allow
users to filter and drill down into the data for more granular
insights.
4. Real-Time Data:
o Power BI supports real-time data feeds, allowing users to
monitor dashboards that update live as new data is received.
5. Advanced Analytics and AI:
o Power BI provides built-in advanced analytics capabilities like
forecasting, trend analysis, and the integration of AI-driven
features like natural language queries.
6. Collaboration and Sharing:
o Power BI makes sharing reports easy through the Power BI
Service (cloud-based) or by embedding reports in apps or
websites.
o The tool allows for team collaboration, sharing, and
commenting on visualizations.
7. Integration with Microsoft Products:
o Power BI is deeply integrated into Microsoft’s ecosystem (e.g.,
Office 365, SharePoint), making it an ideal choice for
organizations already using Microsoft tools.
8. Custom Visuals:
o Users can add custom visuals through Power BI’s marketplace,
or they can create their own using D3.js or R scripts.
9. Cost:
o Power BI offers a free version for individual users, but the paid
version, Power BI Pro, is required for advanced features like
sharing and collaboration. Power BI Premium offers enterprise-
level capabilities.

Best For:

 Organizations already using Microsoft tools.

 Users who prefer a budget-friendly and scalable tool.
 Real-time and cloud-based reporting.

2. Tableau

Overview:
Tableau is a powerful data visualization tool that allows users to create
complex and beautiful visualizations. It is widely used in industries for
business intelligence, data analysis, and storytelling with data. Tableau is
known for its flexibility and ability to handle large datasets with ease.

Key Features:

1. Data Connection and Integration:

o Tableau supports a wide range of data connectors, including
databases, spreadsheets, cloud services, and even web-based
data. It can connect to real-time data sources and large data
warehouses.
2. Drag-and-Drop Interface:
o Similar to Power BI, Tableau offers a drag-and-drop interface to
build visualizations, making it easy to create reports and
dashboards without requiring programming skills.
3. Advanced Data Analytics:
o Tableau allows users to perform complex calculations, create
calculated fields, and perform statistical analysis. It also
provides more flexibility in data preparation and transformation
compared to Power BI.
4. Visualizations and Dashboards:
o Tableau is known for its rich, interactive visualizations,
including bar charts, line graphs, scatter plots, maps, heatmaps,
and more. It also has built-in tools to create powerful
dashboards with drill-down features.
5. Data Blending and Pivoting:
o Tableau allows you to blend data from multiple sources into a
single visualization and pivot tables to easily summarize
complex datasets.
6. Tableau Server and Tableau Online:
o With Tableau Server (on-premise) or Tableau Online (cloud-
based), users can share and collaborate on reports and
dashboards securely. Tableau Server offers more flexibility in
large-scale deployments.
7. Storytelling:
o Tableau allows users to create data-driven stories that guide the
viewer through the analysis with interactive features. This
makes Tableau ideal for presenting data in a narrative format.
8. Mobile Support:
 oTableau has mobile apps (iOS and Android) that allow users to
access visualizations and dashboards on the go, ensuring that
decision-makers can monitor performance from anywhere.
9. Community and Resources:
o Tableau has a large, active community with plenty of resources,
tutorials, and user forums. It also provides a marketplace for
custom visuals.
10.Cost:
o Tableau offers a free version called Tableau Public, but it has
limited features. Paid versions like Tableau Desktop, Tableau
Server, and Tableau Online require a subscription, with pricing
depending on the number of users and deployment type.

Best For:

 Users who need high customization and advanced features.

 Organizations that handle complex, large datasets.
 Those who want powerful data visualization with deep analytical
capabilities.

Key Differences Between Power BI and Tableau:

Feature Power BI Tableau

Simple and intuitive, More complex, but very
Ease of Use
good for beginners flexible and powerful
Excellent for Microsoft Great for large data
Data
tools, good for many sources and varied
Integration
sources formats
Offers a wide variety, Extremely
Visualization
but not as flexible as customizable, high-
Options
Tableau quality visualizations
Free version with
Free version available,
Cost limitations, higher cost
Pro is cost-effective
overall
Excellent support for Supports real-time, but
Real-Time Data
real-time data can be more complex
Advanced AI features, good Strong in advanced
 Analytics integration with Excel analytics and data
manipulation
Strong in Microsoft Strong collaborative
Collaboration ecosystem and cloud features with Tableau
sharing Server/Online
Tableau Mobile app for
Mobile Support Power BI app for mobile
tablets and phones

Which Tool to Choose?

1. Choose Power BI if:

o You are already using Microsoft products (Excel, Azure, etc.).
o You need a cost-effective solution for smaller teams or
departments.
o You need simple visualizations and easy-to-use features for
business analytics.
o You prefer a cloud-based solution with easy sharing
capabilities.
2. Choose Tableau if:
o You need high customization and advanced analytics.
o You work with large and complex datasets and need a tool that
can handle them efficiently.
o You need top-tier visualizations and storytelling features.
o Your team requires a highly flexible tool with extensive
capabilities in data preparation and integration.