Visuals or pictures convey meaning more easily than thousands or even millions of data points. Visualization also allows analysts and end users to recognize patterns and relationships in large volumes of data that may not be easily seen in the raw data or reports. This may help identify emerging trends, for example, to allow an organization to address quality or safety issues before they become bigger problems. The goal is to provide actionable insights that help drive change.
How it can help?
· Identify areas that need attention or improvement
· Clarify which factors influence customer behavior
· Help you understand which products to place where
· Predict sales volumes, Customer behavior and many more….
Use case of data visualization:
· Healthcare: Helping the patients to diagnose early and prevent from disease by seeing patterns from past medical records
· Insurance: Fraud detection and potential customers/areas to focus
· Automotive: Finding the trend and predict the future sales , demand and supply
· Telecom: Customer retention and churn prediction
· Banking: From credit risk to loan and from retail to corporate can predict the defaulters and high paying customers
· Social media, E-Commerce etc.. regardless of industry or size, all types of businesses are using data visualization to help make sense of their data
How charts plays vital role in visualization.
A chart to visualization is as similar to a soul to body. It is well proven that a sign/chart/design conveys more strong messages than a paragraph of text, but when used appropriately.
Let’s find out when to use which chart, what role each chart play and why are they important.
Deviation
When you want to show data in (positive or negative) pattern. Generally the reference point is zero or can be some threshold value.
Diverging bar chart
A simple bar graph where we can show +ve and –ve values.
Diverging Stacked bar
Good way to represent diverging data with average values e.g (survey result – Agree/Disagree/Neutral)
Spine Chart
We can show the % composition of 2 values e.g (True/False, Hot/Cold, Male/Female)
Line chart
A line chart with green and red color indicates +ve and –ve values. e.g profit ratio across time.
Correlation
Shows the relationship between two or more variables. How one variable is dependent or independent to other. E.g Sales vs profit, market share vs Cagr.
Scatter plot
Use to display values for typically two variables for a set of data. Divide data in quadrants and focus on good and poor performing area.
Combo chart
Combo chart combine more than one value. Generally use to see the sales and % growth.
Bubble chart
A bubble chart is a variation of a scatter chart in which the data points are replaced with bubbles, and an additional dimension of the data is represented in the size of the bubbles.
Heat map
Heat Map is two dimensional representation of data in which values are represented by colors. Larger the value more correlated the dimensions are.
Ranking
A ranking is a relationship between a set of items such that, for any two items, the first is either higher, lower or equal to the other.
Ordered Bar
Standard bar charts display the ranks of values much more easily when sorted into order.
Use when there are more number of records.
Ordered column
Standard bar charts display the rank of values much more easier when sorted into order.
Use when there are less number of records.
Dot strip Plot
Dots placed in order on a strip are a space-efficient method of laying out ranks across
multiple categories.
Lollipop chart
It is identical to bar chart but visually it pays more attention to data then a bar chart. It can show rank more effectively.
Slope Chart
Good way of showing how rank have change during last and this year between different categories.
Bump Chart
Another way of showing the rank have changed over a period of time between categories.
Distribution
Distribution charts are used to see how quantitative values are distributed along an axis from lowest to highest. Distributions charts allow users to identify characteristics such as the range of values, central tendency, shape and outliers.
Histogram
A histogram is a bar graph of raw data that creates a picture of the data distribution. The bars represent the frequency of occurrence by classes of data. A histogram shows basic information about the data set, such as central location, width of spread and shape.
Box plot
A box plot is a standard way of displaying the distribution of data five number summary (‘minimum’, ’first quartile’, ‘median’, ’third quartile’, ‘maximum’) It can tell about outliers and what their values are.
Violin plot
Similar to a box plot but more effective with complex distributions (data that cannot be summarized with simple average).
Population pyramid
A standard way for showing the age and sex breakdown of a population distribution, effectively.
Dot plot
A simple way of showing the change or range (min/max) of data across multiple categories.
Barcode plot
Like dot plot, good for displaying all data in a table. They work best when data is highlighted.
Change over Time
When you data is dependent on time period for e.g sales over time , stock price etc. Time period may varies in seconds, minutes, hours, days or years. Choosing a right period gives the better results.
Line chart
The standard way to show a changing time series. If data are irregular, consider markers to
represent data points.
Slope
Good for showing changing data as long as the data can be simplified into 2 or 3 points without missing a key part of story.
Column
Columns work well for showing change over time - but usually best with only one series of data at a time.
Area
It is based on line chart. The area between axis and line are commonly emphasized with colors, textures and hatching. Commonly one compares two or more quantities with an area chart.
Vertical Line
Simple chart use to show the event at a particular date.(year/quarter/month).
Calendar Heatmap
A great way of showing temporal patterns (daily, weekly, monthly) – at the expense of showing precision in quantity.
Stock chart
Stock chart gives you the in depth look at thousand of stocks. You can change the appearance of the chart by varying the time scale. Usually focus on day to day activity.
Forecasting
In general this is use to estimate the forecasting data based on past pattern and seasonality.
Part to whole
Show the contribution of each elements/item with respect to total values. All the elements sum up to 100%.
Stacked column
Simple way to show the 100% contribution it may be easy to read with few items.
Proportional Stacked Bar
Way of showing the size and proportionate of data. Easy to read with few items.
Pie
Very common chart to show part to whole. But difficult to read when items have nearly same values.
Donut
Similar to pie chart. We can show the total/aggregated value in center.
Tree Map
Use for hierarchical part-to-whole relationships can be difficult to read when there are many small segments.
Sunburst
Another way of visualizing hierarchical part-to- whole relationships. Use sparingly (if at all) for obvious reasons.
Arc
Very similar to donut chart.100% corresponds to upper part from left to right.
Gridplot/matrix plot
Matrix chart is a table made up of rows and columns. Easily can show the % value on it.
Waterfall
Can be useful for showing part-to-whole relationships where some of the components are negative.
Pictogram
Pictogram Charts use icons to give a more engaging. Overall view of small sets of discrete data. Number of pictures gives the % number. Generally used in population data (Male/Female)
Magnitude
Show size comparisons. These can be relative (just being able to see larger/bigger) or absolute (need to see fine differences). Usually these show a ‘counted’ number (for example, barrels, dollars or people) rather than a calculated rate or percent.
Paired Column
A simple column chart made on 2 dimensions. Can be difficult to read more than two. For e.g time and region.
Paired Bar
As per standard bar, but allows for multiple series. Can become tricky to read with more than 2 series.
Spider/Radar Chart
A space-efficient way of showing value of multiple variables– but make sure they are organized in a way that makes sense to reader.
Parallel Coordinates
An alternative to radar charts – again, the arrangement of the variables is important. Usually benefits from highlighting values.
Flow
Show the reader volumes or intensity of movement between two or more states or conditions. These might be logical sequences or geographical locations.
Sankey
Shows changes in flows from one condition to at least one other; good for tracing the eventual outcome of a complex process.
Waterfall
Designed to show the sequencing of data through a flow process, typically budgets. Can include +/- components.
Chord
A complex but powerful diagram which can illustrate 2-way flows (and net winner) in a matrix.
Network
Used for showing the strength and inter-connectedness of relationships of varying types.
Spatial
Used only when precise locations or geographical patterns in data are more important to the reader than anything else.
Flow Map
For showing unambiguous movement across a map. Generally used when we have source and destination.
Heat Map
Grid-based data values mapped with an intensity colour scale. As choropleth map – but not snapped to an admin/political unit.
Dot Density
Used to show the location of individual events/locations – make sure to annotate any patterns the reader should see.
Magnitude Map
Use for totals rather than rates – be wary that small differences in data will be hard to see.
Note:
The content is inspired from Chart guide and Visual Vocabulary author Andy Kriebel, Originally published on Tableau Public.
Many charts mentioned above are easy to make in tableau. Please refer Tableaureferenceguide.com to know more in detail.
An Article by Rohan Makhija
Rohan Makhija has 5+ years of work experience with Tableau Desktop.Working as Data Analyst in Management consulting where role is to find the solution to business problem by analyzing the data with the tools like Tableau, SQL, Alteryx to drive data into meaningful insights and present into storyline to the business unit to take better decisions. Graduated in Computer Science from Kurukshetra University and Post-Graduation Diploma in Data Analysis from IIIT Bangalore.
Linked In: https://www.linkedin.com/in/rohanmakhija05/
Comments