Insights for ArcGISDavid Thom
Solution Engineer – State Government
ArcGIS Apps Bring the Power of Location to EveryoneExtending the Reach of GIS
Dashboards
Story Maps
Social MediaViewer AppsDesign Apps
Business Apps
Field Apps
Across Organizations and Beyond
ArcGIS | Common Patterns of Use
ArcGIS | Common Patterns of Use
Art Haddad, CTO, Location Analytics
“Data is important…
However, not as important as the answers it
provides or the stories that it tells.”
To understand the data, one must ask questions…
Where can you get a competitive advantage?
Which customers are in this area?
What Factors are Affecting Profitability in NYC vs. LA ?
What’s the Total Insured Value of Clients in this Area?
Where should I be doing better?
Where is your business most exposed?
Where are people most in need?
The Process
of Analysis
Steps of Analysis
1. Begins with a well-framed
question
2. Manipulate, quantify, and
manage your data
Ask and
explore1
Model and
compute2
Examine and
interpret3
Share
results5
Make
decisions4
1. Increasingly data is being
automatically collected
2. The analyst must find the
value in the collected data
An analyst must understand the data, analysis and the intersection between the two
Available with Online or Enterprise
Your infrastructure(physical, virtual, or cloud)
ArcGIS Enterprise
ArcGIS Online
OR
Analytic workflows
ArcGIS Insights
Visualization
• Interactive data views
• Charts, graphs, tables, and maps
• Dynamic selection & cross filtering
• Context of the largest digital atlas in world
Data Preparation
• Format fields
• Calculate new fields
GeoEnablement
• Geocode address
• Convert X,Y
• Join to boundaries
• Join to demographic variables
• Descriptive stats
• Link analysis
• Regression
Statistical AnalysisSpatial Analysis
• How is it distributed?
• How is it related?
• What’s nearby?
• How has it changed?
Record & Repeat Analysis
• Share model to analysts
• Document best practices
• Re-run with new data
Share Results
• Share with executives
• Tell your story
• Share with public
Transforming Understanding with Interactive Analysis
Extending and Complementing Traditional Analysis…
. . . Supporting a System of Insight
Insights
• New Paradigm (Visual, Intuitive, Responsive)
• Dynamic Analysis
• Very Fast Data Access
On-The-Fly Visual Models
Analytic Projects – Workbook, Pages, Cards
What data does Insights work with?
• Feature Layers (WebGIS)
• Spreadsheets
• GeoDatabase
• Business Databases
• Big Data
• Real-time data
Adding Data to Insights
• When adding data to a workbook, each layer, table, or view is added as a separate dataset by
default.
Get started guide
Categorical
Numeric
Shape field
Date/time
Date
components
Data pane
Dataset
Result
dataset
Hierarchy
Indicates
card-dataset
connections
Results of Analytic Operations are expressed in Cards
Visualizations
Distribution: the arrangement of phenomena, could be numerically or spatially
Measure: ascertain the size, amount, or degree of (something)
Change: process through which something becomes different, often over time
A Data clock creates a circular chart of temporal data,commonly used to see the number of events at differentperiods of time.
A box plot displays data distribution showing the median,upper and lower quartiles, min and max values and, outliers.Distributions between many groups can be compared.
Histograms show the distribution of a numeric variable.The bar represents the range of the class bin with theheight showing the number of data points in the class bin.
A heat chart shows total frequency in a matrix. Using atemporal axis values, each cell of the rectangular grid aresymbolized into classes over time.
Line graphs visualize a sequence of continuous numericvalues and are used primarily for trends over time. Theyshow overall trends and changes from one value to the next.
A bar graph uses either horizontal or vertical bars to showcomparisons among categories. They are valuable toidentify broad differences between categories at a glance.
Bubble charts represent numerical values of variables byarea. With two variables (category and numeric), the circlesplaced so they are packed together.
A bar graph uses either horizontal or vertical bars to showcomparisons among categories. They are valuable toidentify broad differences between categories at a glance.
A treemap shows both the hierarchical data as a proportionof a whole and, the structure of data. The proportion ofcategories can easily be compared by their size.
A heat chart shows total frequency in a matrix. Values ineach cell of the rectangular grid are symbolized intoclasses.
Graduated symbol maps show a quantitative differencebetween mapped features by varying symbol size. Data areclassified with a symbol assigned to each range.
A choropleth map allows quantitative values to be mappedby area. They should show normalized values not countscollected over unequal areas or populations.
A chord diagram visualizes the inter-relationships betweencategories and allows comparison of similarities within adataset or, between different groups of data.
Scatterplots allow you to look at relationships between twonumeric variables with both scales showing quantitativevariables. The level of correlation can also be quantified.
A Density/heat map calculates spatial concentrations ofevents or values enabling the distribution to be visualizedas a continuous surface.
Interaction: flow of information, products or goods between places
A chord diagram visualizes the inter-relationships between categories and allows comparison of similarities within a dataset or, between different groups of data.
A combo chart combines two graphs where they sharecommon information on the x-axis. They allow relationshipsbetween two datasets to be shown.
A Density/heat map calculates spatial concentrations ofevents or values enabling the distribution to be visualizedas a continuous surface.
Graduated symbol maps show a quantitative differencebetween mapped features by varying symbol size. Data areclassified with a symbol assigned to each range.
A unique symbol map (areas or points) allows descriptive(qualitative) information to be shown by location. Areashave different fills and points can be geometric or pictorial.
A choropleth map allows quantitative values to be mappedby area. They should show normalized values not countscollected over unequal areas or populations.
Spider lines , also termed desire lines, show paths betweenorigins and destinations. They show connections betweenplaces.
Spider lines, also termed desire lines, show paths betweenorigins and destinations. Flow maps show directionalconnections and flow between places.
Relationship: a connection or similarity between two or more things or, the state of being related to something else
Donut charts are used to show the proportions ofcategorical data, with the size of each piece representingthe proportion of each category.
A treemap shows both the hierarchical data as a proportionof a whole and, the structure of data. The proportion ofcategories can easily be compared by their size.
AcknowledgementInspired by work by Jon Schwabish and SeverinoRibecca, The Graphic Continuum, 2014 and, Alan Smith et al. Visual Vocabulary, The Financial Times, 2016
QuantitativeQualitative TemporalData type:
Linda Beale PhD, 2017
Bubble charts with three numeric variables are multivariatecharts that show the relationship between two values whilea third value is shown by the circle area.
Part-to-whole: relative proportions or percentages of categories, showing the relationship between parts and whole
Link analysis is used to investigate relationships betweenentities where and an entity is an object, person, place orevent. Links connect two or more entities.
Analysis: starts with questions
Analysis: Interactivity or Action Button
Analysis: Interactivity
What Analytic Operations can I do in Insights?
Queries
Filters
Summary Tables
Attribute Joins
Field Calculations
A B
Data Analysis:
Field Calculations Summary StatsΣEnrichment
Plus:
Spatial Filter
Spatial
Aggregation
Fixed Distance & Drive-time buffers
Kernel Density
Find Nearby
Find Similar
Spatial Analysis:
Spatial Join
Relationships
Quantify results: Statistical reporting
Working with tables
Data table
Table card
A window into the steps of analysis
Sharing in Insights
Data
Page
Model
Workbook
David Thom
Let’s See It
In Summary…
• Exploratory and Interactive Spatial Analysis
- Core capability in ArcGIS
• Focus on GIS
- Go beyond basic mapping…
- Harness the full power of ArcGIS and Spatial Analysis
• Start with the Questions, get the Answers, Document the workflows, and Share your stories.
NEW FEATURES
Advanced filters
Calculate field
Change unique
colors
Public sharing
(Online)
Analytic Journals
• Page and card options
• Widgets
- e.g. rich text card
Link Analysis +
• Used to investigate relationships between entities
- where and entity is an object, person, place or event
- A link connects two or more entities
• Link charts
• Spider lines/Desire lines
• Flow maps