Analyzing Data Using | My Assignment Tutor

Copyright © SAS Institute Inc. All rights reserved.Lesson 3: Analyzing Data UsingSAS® Visual Analytics 3.1 Working with Data Items3.2 Exploring Data with Charts and Graphs3.3 Creating Data Items and Applying Filters3.4 Performing Data Analysis Copyright © SAS Institute Inc. All rights reserved.Lesson 3: Analyzing Data UsingSAS® Visual Analytics 3.1 Working with Data Items3.2 Exploring Data with Charts and Graphs3.3 Creating Data Items and Applying Filters3.4 Performing Data Analysis Copyright © SAS Institute Inc. All rights reserved.3Objectives• Discuss the Analyze phase of the SAS Visual Analytics methodology.• Change data items (modify formats, modify aggregations, modifyclassifications, rename data items) in Visual Analytics for the analysis.Copyright © SAS Institute Inc. All rights reserved.4Visual Analytics Methodology: AnalyzeIn the Analyze phase, you canevaluate the data by doing thefollowing:• modifying data item properties• creating new calculated items needed for analysis• applying any necessary filters for the analysis• exploring relationships between data items using charts and graphs• discovering trends and patterns between data items• creating, testing, and comparing models based on patterns discoveredAccess Investigate Prepare Analyze ReportCopyright © SAS Institute Inc. All rights reserved.5SAS Graph BuilderSAS Viya ApplicationsSAS Cloud Analytic Services(CAS)SAS Report ViewerSAS Data StudioSAS Theme DesignerSAS Visual AnalyticsSAS Visual Analytics AppSAS DriveCopyright © SAS Institute Inc. All rights reserved.6Business Scenario: CustomersBased on the investigation of the data and the assignment (analyze profitsfor the Marketing team and analyze delivery times for the Shipping team),you need to make some changes to data items in the CUSTOMERS table.ModifyFormatsAggregationsRename Data Items68,300 customers 747,953 ordersYou can make more changes asyou perform the analysis.Copyright © SAS Institute Inc. All rights reserved.7SAS Data Studio versus Visual AnalyticsSAS Data Studio Visual AnalyticsCloud Analytic Services(CAS)Report Viewer Visual Analytics AppData ViewCopyright © SAS Institute Inc. All rights reserved.8Data Item PropertiesIn the Data pane, properties can be modified for each data item to aidin your analysis.Category DatetimeMeasureCopyright © SAS Institute Inc. All rights reserved.Working with Data ItemsThis demonstration illustrates how tomodify data item properties (name,format, aggregation) in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.10Business Scenario: EmployeesBased on your investigation of the data and your assignment (analyzesalaries to determine employees who can be promoted), you need to makesome changes to data items in the EMPLOYEES table.ModifyClassificationsFormatsYou can make more changes as Rename Data Itemsyou perform the analysis.648 employeesCopyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.Lesson 3: Analyzing Data UsingSAS® Visual Analytics 3.1 Working with Data Items3.2 Exploring Data with Charts and Graphs3.3 Creating Data Items and Applying Filters3.4 Performing Data Analysis Copyright © SAS Institute Inc. All rights reserved.14Objectives• Discuss when to use descriptive graphs (histogram, box plot, bar chart)in Visual Analytics.• Maximize graphs objects to view details.• Modify roles and options for graph objects.Copyright © SAS Institute Inc. All rights reserved.15Business Scenario: CustomersFor the Marketing team, you have been asked to analyze profits. As a firststep, you would like to understand the range of profits generated by OrionStar, as well as total profits for different order types and from differentcontinents.You will then use this analysis to determine thefocus group for the next marketing campaign.Copyright © SAS Institute Inc. All rights reserved.16Objects: Graphs (Descriptive)Use a histogram to view thedistribution of a single measure.Use a box plot to view informationabout the variability of the data,extreme values and outliers.outliers are data points whose distance fromthe interquartile range are more than 1.5times the size of the interquartile rangeCopyright © SAS Institute Inc. All rights reserved.17DistributionsP (X)0.4 –0.3 –0.2 –0.1 –| | | | | |1 2 3 4 5X• Central tendency ofthe distribution is the•is th mean/average/expected valueAmount of variabilityvarianceCopyright © SAS Institute Inc. All rights reserved.18Common distributionsUniform NormalCentred around meanand symmetricCopyright © SAS Institute Inc. All rights reserved.19Common distributionsBimodal Lognormal2 peaks right skewed-> 2 different groups-> can we separate them?Copyright © SAS Institute Inc. All rights reserved.20Normal distributionmSame m, smaller sSame m, larger sm = means = standard deviationCopyright © SAS Institute Inc. All rights reserved.21Box plotAlso known as Whiskers plotsCopyright © SAS Institute Inc. All rights reserved.22Box plotDealing with outliersCopyright © SAS Institute Inc. All rights reserved.23Box plot: How ToAge: 18 18 18 19 19 19 20 20 20 20 20 21 21 21 21 21 22 22 22 50Sorted observationsCopyright © SAS Institute Inc. All rights reserved.24Box plot: How ToAge: 18 18 18 19 19 19 20 20 20 20 20 21 21 21 21 21 22 22 22 5025% of observations ≤ 191st quartileCopyright © SAS Institute Inc. All rights reserved.25Box plot: How ToAge: 18 18 18 19 19 19 20 20 20 20 20 21 21 21 21 21 22 22 22 5050% of observations ≤ 201st quartile or medianCopyright © SAS Institute Inc. All rights reserved.26Box plot: How ToAge: 18 18 18 19 19 19 20 20 20 20 20 21 21 21 21 21 22 22 22 5075% of observations ≤ 203rd quartileCopyright © SAS Institute Inc. All rights reserved.27Box plot: How ToAge: 18 18 18 19 19 19 20 20 20 20 20 21 21 21 21 21 22 22 22 50• 1st quartile (Q1): 19• Median: 20• 3rd quartile (Q3): 21• IQR: 21-19 = 2• Min: 18• Max: 50• Min*: Q1 – 1.5 x 2 = 16 (< Min => use actual Min)• Max*: Q3 + 1.5 x 2 = 24• Average: 21.6Q1Q3medianminmaxoutlieraverageCopyright © SAS Institute Inc. All rights reserved.28Discussion: Median vs AverageWhy using median instead of average?Copyright © SAS Institute Inc. All rights reserved.293.01 Multiple Choice QuestionWhich graph would help you determine whether a measure is normallydistributed?a. box plotb. histogramc. scatter (X/Y) plotCopyright © SAS Institute Inc. All rights reserved.303.01 Multiple Choice QuestionWhich graph would help you determine whether a measure is normallydistributed?a. box plotb. histogramc. scatter (X/Y) plotCopyright © SAS Institute Inc. All rights reserved.33Objects: Graphs (Descriptive)Use a bar chart to compare summarized data for the following:Nominal valuesRankingsTime series dataParts of a wholeDisplays dataaggregated bycategoriesCopyright © SAS Institute Inc. All rights reserved.Exploring Data: Part 1This demonstration illustrates how touse the automatic chart to exploredata and modify roles and options forcharts and graphs in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.35Business Scenario: EmployeesFor the Human Resources team, you have been asked to analyze salaries todetermine which employees could be eligible for promotion. As a first step,you would like to understand the range of salaries at Orion Star, as well astotal salaries by job title.You will then use this analysis to determine the employees targeted forpromotion.Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.39Business Scenario: CustomersIn the previous analysis, you discovered that profits were lower in theinternet and catalog channels. Continue to analyze profits by order typeto determine ways to improve profits through these channels.You also discovered that profits were lower inNorth America than in Europe, even though youexpected the opposite. Continue to analyzeprofits by location to understand why thisdiscrepancy exists and determine ways toimprove profits in non-European countries.Copyright © SAS Institute Inc. All rights reserved.Exploring Data: Part 2This demonstration illustrates how to usebox plots to explore data in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.41Business Scenario: EmployeesIn the previous analysis, you discovered that salary costs were higher foremployees with the Sales Rep. I title. Continue to analyze salary costsby job title to determine employees that might qualify for promotion.Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.Lesson 3: Analyzing Data UsingSAS® Visual Analytics 3.1 Working with Data Items3.2 Exploring Data with Charts and Graphs3.3 Creating Data Items and Applying Filters3.4 Performing Data Analysis Copyright © SAS Institute Inc. All rights reserved.45Objectives• Describe the types of data items that can be created in Visual Analytics.• Discuss the difference between calculated items and aggregatedmeasures.• Describe the various ways that data can be filtered in Visual Analytics.• Discuss when to use geographic maps in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.46Business Scenario: CustomersIn the previous analysis, you discovered higher total profits for retailsales despite slightly higher average profits for internet and catalog sales.Why are the total profits higher for this group?In addition to the analysis of profits by order typeand continent, you also need to analyze profits bygender and age group to determine a focus groupfor our next marketing campaign.You need to create new dataitems for this analysis.Copyright © SAS Institute Inc. All rights reserved.47Creating Data ItemsThe following data items can be created in Visual Analytics:Calculated GeographyHierarchy Custom category Derived itemsDuplicate Distinct count15ParametersCopyright © SAS Institute Inc. All rights reserved.48Why Duplicate? Order Total1818181919192020202020212121212122222250GenderMMMFMFFMFMMFFMFMFFMF Default aggregation: Sum• Order Total for M: 197• Order Total for F: 235Duplicate the variable allows to change the aggregation type, eg to Average:• Order Average for M: 19.7• Order Average for F: 23.5Can use the 2 aggregations in the same report!Copyright © SAS Institute Inc. All rights reserved.49Calculated Item: ExampleCalculated items are created by performing operations on unaggregateddata. GenderSalaryIncreaseNew SalaryMale40,0001.0542,000Female65,0001.1071,500Female32,0001.0533,600Male80,0001.1088,000Female56,0001.1564,400 Copyright © SAS Institute Inc. All rights reserved.50Aggregated Measure: ExampleAggregated measures are created by aggregating and then performingthe operation. GenderSalaryMale120,000Female153,000TOTAL273,000 GenderSalaryMale40,000Female65,000Female32,000Male80,000Female56,000 Copyright © SAS Institute Inc. All rights reserved.513.02 ActivityMatch each new data item with the type of calculation.___ Gross Profit Margin (Total Profit/ Total Revenue)___ Date (from month, day, year)___ Hemisphere (from continents)___ Number of Employees (distinct count)___ State Abbreviations (uppercase)A. calculated itemB. aggregated measureCopyright © SAS Institute Inc. All rights reserved.53Custom Category: ExampleCustom categories create labels for groups of category or measure dataitems.This calculated item and custom categoryproduce equivalent results.Calculated item Custom categoryCopyright © SAS Institute Inc. All rights reserved.543.03 ActivityGiven the values of Customer Birth Date and today’s date, how would youcalculate Customer Age?Copyright © SAS Institute Inc. All rights reserved.553.03 Activity – Correct AnswerGiven the values of Customer Birth Date and today’s date, how would youcalculate Customer Age?Customer Age = (Today – Customer Birth Date)/365.25Note:In SAS, dates are stored as the number of days since January 1, 1960Copyright © SAS Institute Inc. All rights reserved.56Calculated Columns: Customer AgeThe Now operator creates a datetime value using the current date and time,where the current date and time is evaluated every time you view thereport (date + time).The DatePart operator converts a datetime value to a date value.The TreatAs operator enables a numeric, or datetime, value to be used as adifferent data type within other operators.The Floor operator rounds the number down to the nearest integer.Copyright © SAS Institute Inc. All rights reserved.57Objects: Graphs (Geography)Use a geo map when location is a critical component of the analysis.Bubbles CoordinatesRegionsUse a geo contourmap to show verydense data.ContourUse a geo region mapor geo coordinate maponly when there is aneven distribution ofvalues within eachregion.Copyright © SAS Institute Inc. All rights reserved.583.05 Multiple Answer QuestionWhich object can use a data item that has a classification typeof geography?a. crosstabb. geo mapc. tabled. bar chartCopyright © SAS Institute Inc. All rights reserved.593.05 Multiple Answer Question – Correct AnswerWhich object can use a data item that has a classification typeof geography?a. crosstabb. geo mapc. tabled. bar chart• All these graphs can use a data item that has a classification typeof geography. The geo map requires it.Copyright © SAS Institute Inc. All rights reserved.60What Is a Hierarchy?MonthQuarterYear 20191JanFebMar2AprMayJun3JulAugSep4OctNovDecParent-child relationshipsCopyright © SAS Institute Inc. All rights reserved.Creating Data ItemsThis demonstration illustrates how tocreate new data items (distinct counts,custom categories) in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.62Business Scenario: EmployeesIn the previous analysis, you discovered higher salary costs for employeeswith the Sales Rep. I title despite having relatively low average salary costs.Why are total salary costs higher for this group?In addition to the analysis of salaries by job title, you also need to analyzethe type of employee (active versus retired) and years of service todetermine which employees to target for the next round of promotions.You need to create new dataitems for this analysis.Copyright © SAS Institute Inc. All rights reserved.633.04 ActivityGiven the values of Employee Hire Date and Employee Termination Date,how would you calculate Years of Service?Copyright © SAS Institute Inc. All rights reserved.643.04 Activity – Correct AnswerGiven the values of Employee Hire Date and Employee Termination Date,how would you calculate Years of Service?You would need two different calculations:one for active employees and one forretired employees.IF Active employees:YOS = (Today – Employee Hire Date)/365.25IF Retired employees:YOS= (Employee Termination Date –Employee Hire Date)/365.25Use the IF… ELSE operator to perform different calculations based on a condition.Retired Active Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.68Business Scenario: CustomersManagement has decided that our initial marketing strategy should focus onincreasing sales among North American customers who order throughcatalog and internet. If successful, we can push the campaign to otherlocations.The Marketing team has asked how profits aredistributed throughout the United States to seewhether there are any clusters that can be identifiedand used for the campaign.You need to create new data items, add a filter,and create a hierarchy for this analysis.Copyright © SAS Institute Inc. All rights reserved.69Filtering DataMany different types of filters can be created to subset data in Visual Analytics:Report Designer Report ViewerPrompts• Report• PageActions• Filter• Links• Data source• Basic• Advanced• Post-aggregate report filtersReport objectsWhole reportCopyright © SAS Institute Inc. All rights reserved.Applying FiltersThis demonstration illustrates how tocreate new data items (geographicdata items, hierarchies) and applyfilters in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.71Business Scenario: EmployeesManagement has decided that your initial promotion analysis should focuson active employees in the Sales Department.The amount of profit generated by each employee has been identified asone possible criterion for promotion. Given this criterion, you need toidentify locations where initial promotions should begin.Create new data items and adda filter for this analysis.Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.Lesson 3: Analyzing Data UsingSAS® Visual Analytics 3.1 Working with Data Items3.2 Exploring Data with Charts and Graphs3.3 Creating Data Items and Applying Filters3.4 Performing Data Analysis Copyright © SAS Institute Inc. All rights reserved.77Objectives• Discuss when to use analysis graphs in Visual Analytics.• Describe the types of fit lines that can be added to analysis graphs.• Describe the forecasting capabilities available in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.78Business Scenario: CustomersThe Shipping team has suggested that delivery times could be responsiblefor the lower profits in the internet and catalog channels. They have askedthat you determine how delivery times, number of orders, and profits arerelated.As you work on this analysis, the Marketingteam has asked for help with determining the focusgroups for the next marketing campaign by analyzingorder types, genders, and age groups.Copyright © SAS Institute Inc. All rights reserved.79Objects: Graphs (Analysis)Use a bubble plot to display threedimensions of data (horizontal location,vertical location, size of bubble) forsome group of category values.Use a treemap to display lots ofinformation in a small amount ofspace. Use size and color to drawattention to specific areas of interest.Copyright © SAS Institute Inc. All rights reserved.Analyzing DataThis demonstration illustrates how toanalyze data with graphs in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.81Business Scenario: EmployeesThe Human Resources team has suggested that employees who have beenwith the company longer and those who have generated higher profitsshould be promoted.The team has asked you to identify the companies and job titles where theyshould begin promotions.Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.Copyright © SAS Institute Inc. All rights reserved.86Business Scenario: CustomersTo complete the analysis, the manager has asked that you analyze therelationship, if any, between delivery times, discounts, total revenue,and unit costs.In addition, you need to determine the relationship between profits andnumber of orders and predict how these trends will continue in the future.Copyright © SAS Institute Inc. All rights reserved.87Correlation • Measure between -1 and 1• Association between 2 variables• Noisiness and direction of linear relationship (but not the slope) Weak:|𝜌| ∈ [0,0.3[Moderate: |𝜌| ∈ [0.0,0.6[Strong:|𝜌| ∈ [0.6,1[ Copyright © SAS Institute Inc. All rights reserved.88CorrelationDirection of theassociation ?Copyright © SAS Institute Inc. All rights reserved.89CorrelationIn this case, it is quite clear:good weather increasessales of ice-creamCopyright © SAS Institute Inc. All rights reserved.90Correlation is not causation!Is ice cream dangerous?Copyright © SAS Institute Inc. All rights reserved.91Correlation is not causationShark attacks, bushfires and ice-cream sales are correlated because they are also correlated with hot weather!Copyright © SAS Institute Inc. All rights reserved.92Objects: Graphs (Analysis)Use a scatter plot to evaluate therelationship between two measures.Use a correlation matrix to evaluate thelinear relationship between measures.Use a heat map to evaluate the relationship betweentwo high-cardinality measures, between twocategories, or between a category and a measure.Copyright © SAS Institute Inc. All rights reserved.95NoteEach report object has a threshold for how much data it can visually display.Many report objects will not display high-cardinality data items with lots ofunique values.Examples of high-cardinality data items:Employee ID, Street Address, Customer Name, Birth DateExamples of low-cardinality data items:Country Name, Age Group, Job Title, Store TypeCopyright © SAS Institute Inc. All rights reserved.96Fit LinesFit lines can be added to scatter plots and heat maps to plot the relationship between variables.LinearQuadratic CubicPSplineCopyright © SAS Institute Inc. All rights reserved.97Be careful!Copyright © SAS Institute Inc. All rights reserved.98Objects: Graphs (Analysis)Use a line chart to show trends oversome ordinal variable (time, age group).Use a time series plot to showtrends of measures over time.Copyright © SAS Institute Inc. All rights reserved.99Visual Analytics automaticallyselects the best forecastingmodel for your data.Objects: Analytics (Forecasting)Use a forecastingobject to showestimates of futurevalues based onhistorical trends inthe data.Default: next 6 periodsCopyright © SAS Institute Inc. All rights reserved.Adding Data AnalysisThis demonstration illustrates how to adddata analysis to graphs in Visual Analytics.Copyright © SAS Institute Inc. All rights reserved.101Business Scenario: EmployeesTo complete the analysis, your manager has asked that you analyzethe relationship, if any, between salary, orders, profit, and years of serviceto determine alternate criteria for promotion.In addition, you need to determine whether there are any job titledifferences between employees identified for promotion basedon the criteria specified by management.Copyright © SAS Institute Inc. All rights reserved.PracticeThis exercise reinforces the conceptsdiscussed previously.

QUALITY: 100% ORIGINAL PAPER – NO PLAGIARISM – CUSTOM PAPER

Leave a Reply

Your email address will not be published. Required fields are marked *