Home

- CiteSeer

1. L 989a 9 Files Files Applications Newstats NewstatsJar RegistrationBean C Applications Entity ed E amp 3 NewstatsWiar 3 NewstatsJar Persistence Management ChiSquareBe Bean Managed Persistence a anta Container managed persistence 1 0 LinearRegres i Container managed persistence 2 0 e i i v seners Regisiratione Fields To Be Persisted Abstract Schema Name Bl localhost zi name UserSchema i password i ine Finder Select Methods 17 loginid 1 sala na N Deployment Settings Primary Key Class java lang String Primary Key Field Name loginid v Reentrant K Click on Finder Select Methods to write EJB OL Click OK after writing the EJB QL 2 Click on Deployment Settings 3 Inthe Deployment Setting dialog box click Database Setting 4 Enter jdbc oracle in the database JNDI name field Enter you user name and password for the Oracle Database server 5 Click OK 6 Click Generate Default SQL 7 Click OK in SQL Generation Complete Inform dialog 8 Click OK Select the Jar file and click on the JNDI tab Add the JNDI names Click OK 137 Deploying The J2EE Application To deploy the application with the WAR and JAR components added select the application name Select Tools Deploy Select the checkbox labeled Return
2. Learring project technologies ddae Wed 49043 Tue 3 EE Develop Executable dye H 1802 Devebpaofiwara quiyseswamos Mon 6 049 EHE Depuis ddl MaB 9 dy m H Phase des sam Archiecture Design didas Fig Ven THANG funeris ctt M t um fefine Ottimi s Wet ids TUS Hays Mon 82503 LE 18 Fomateticanepectin ii uH Refine USE model Mon 928408 Te tO H Refine project Plan le Wed 100003 Thu H Wak cn grddype 13dys Wed 40183 Suv prre oly Won 102748 ma 18 E Implementtin Tue 020 lenem 15elaye Tue 102808 5 B 5 um EB TT Sohys Tue 18108 Tu i 1 iter Rept Seys Monta 3 H Refine project plan day Mon 14 2408 i Figure 5 Gantt chart 26 Cost Estimation In order to achieve efficient and effective management of software projects it is important to estimate the size and cost of the project For the size and cost estimation of my MSE project I will use Function Point Analys
3. expected Total 5 5 10 126 Degrees of freedom 1 Chi Sguare 0 4 Critical Value CV for 5 level of significance 3 841 If Chi Square value is greater than CV than reject the null hypothesis If not than we fail to reject the null hypothesis Figure 21 Table of Chi Square Results Testing using JMeter The following tests were performed by using Jakarta JMeter on the questions dynamic page in the website The number of users was increased and corresponding performance of the website was noted at a loop count of 10 10 30 Users second 45 Users second Users second average worst optimal Throughput 755 min 981 min 824 min 48 The throughput is very much dependent on the network traffic and the overall load on the server It may help the throughput of the application if the database server is completely dedicated to the application The two important parameters in the above table are deviation and average ms If the standard deviation is high some users will experience very good responses while some other users will wait for a longer time The smaller the value the better it is The average value signifies the average amount of time taken to load a particular page For 10 users 127 Apache JMeter File Edit Run Options Help 9 Test Plan 9 E questionsthreadgroup Graph Results HTTP Request Defaults Name Graph Results ff HTTP Request Write All Data to a File S
4. Phase I Duration Documentation Research 990 840 22 26 El Research m Design A 240 Sanny Deploying amp O Deploying amp Testing Coding 6 Testing 1110 18 m Documentation 28 Figure 22 Phase I duration Phase II Duration 151 Documentation Research 390 640 17 10 Design 1120 Deploying amp 29 Testing 810 2196 Coding 900 23 Figure 23 Phase II duration Phase III Duration Documentation MOSS PAM 920 23 3 Design 80 296 Coding 1050 Testing 1800 2596 46 Deploying amp Figure 24 Phase III duration 152 El Research Design D Coding Deploying amp Testing m Documentation G Research m Design Coding O Deploying amp Testing m Documentation Lessons learnt Methodology On completion of the project the developer realizes the usefulness of following the software methodologies and the life cycle The MSE portfolio served as a useful guide throughout the project The developer believes that this project experience has eguipped her with a better understanding of a software life cycle and will be a guiding factor in future projects Reviews One important lesson learnt in doing this project was the importance of reviews and feedback Besides getting valuable feedback from my major advisor the committee as a whole also reviewed my progress and gave valuable inputs during presentations which made this project lea
5. The Hypothesis test will successfully display the correct results e TheChi Sguare test will successfully display the correct results Other documents Divyagyana Nanjaiah and Lakshmikanth Ganti stating that they have successfully participated in the formal inspection of the reguirement specification and thus passed the software reguirements specification will provide two formal technical inspection letters References will be provided for all notations used in the portfolio 36 CHAPTER 3 ARCHITECTURE DESIGN The design of SAT is summarized using an object model as given below It consists of four session beans one entity bean and one servlet Arrows and multiplicities show the relationship between the classes The model consists of a servlet class four session bean classes and an entity bean class The servlet class creates instances of the session beans whenever a member selects a particular test to be performed The tool performs four kinds of analysis namely simple linear regression correlation analysis hypothesis testing and the chi square test The design is developed in such a way that each test is a session bean This design enables the expansion of the tool If a new test other than the above four tests has to be added to the tool the developer has to just write another session bean for it without disturbing the rest of the design Another advantage of the design is that it allows addition of more research questions fo
6. The J2EE architecture with EJB provides the latest development method to implement this tool as it uses multi tier architecture and provides methods to connect the tools and software together Besides the project uses each analysis test as a session bean hence the expansion of the project is simplified by just adding a new separate session bean for a new test The number of research questions for these tests can be added easily into the project as the infrastructure of the tests is already present and a simple modification in the query to obtain relevant data from the database will serve the purpose Thus using EJB with J2EE architecture provides an appropriate technology for the implementation of this project Java 2 Enterprise Edition J2EE Architecture is defined as a basis to define the complete system its structure its interfaces and intercommunication between the components of the system The J2EE architecture is multi tiered and will basically be comprised of the following three main layers e Presentation layer Front Layer Web containers that host presentation and possibly business logic components e Business layer Middle Layer Application containers that host business logic components e Backend layer Data layer Databases RDBMS of any kind which will host the data and the data access methods J2EE applications are made up of components which are self contained functional units that interact with other
7. setJoinIn3Mths public abstract void setJoinIn3Mths boolean join 66 ejbCreate public java lang String ejbCreate java lang String LId java lang String Passwd java lang String name double gpa double salary char degree boolean citizen boolean join throws javax ejb CreateException This method created the user and sets the values for all the fields Parameters login id for the user Passwd password for the user name full name of the user gpa final cumulative gpa for the user salary current salary for the user degree degree for the user MS or PHD join Yes if he joined a job in 3 months after graduation otherwise No 67 ejbPostCreate public void ejbPostCreate java lang String LId java lang String Passwd java lang String Name double salary double gpa char degree boolean citizen boolean join throws javax ejb CreateException setEntityContext public void setEntityContext EntityContext ctx unsetEntityContext public void unsetEntityContext ejbRemove public void ejbRemove ejbLoad public void ejbLoad 68 ejbStore public void ejbStore ejbPassivate public void ejbPassivate ejbActivate public void ejbActivate Class LinearRegressionBean public class LinearRegressionBean LinearRegressionBean is a session bean which handles the computation for the correlation test It tests to see if higher GPA results in higher salaries for the Masters
8. R CorrelationBean HypothesisBean LinearRegressionBean Figure 11 Servlet interaction The detailed method description for this class is as follows init public void init ServletConfig config throws ServletException This method will be called by the Servlet container to initialize the servlet This method created a new context for the session Parameters config the ServletConfig object that contains configuration information for this servlet Throws ServletException if an exception occurs that interrupts the servlet s operation 51 service public void service HttpServletReguest reguest HttpServletResponse response throws java io IOException This method will be called by the Servlet container when this servlet is being placed into service Parameters request the HttpServletRequest object response the HttpServletResponse object Throws java io IOException if an IO exception occurs servlet s operation is interrupted addLine public static void addLine java lang String message java io PrintWriter out throws java io IOException This method adds a line of HTML output to the page Parameters message HTML line to be added out PrintWriter object 52 makeSalaryString public java lang String makeSalaryString double x int n This method collects the salaries of members in an array Parameters x Array for salary n number of items in the array R
9. Using the above formula for the application programs The programmer productivity and the development time are as follows KDSI 3 6 KLOC PM 2 4 3 6 5 9 person month 32 TDEV 2 5 9 8 5 75 month Architecture Elaboration plan The activities and actions to be accomplished prior to the architecture presentation are listed below The documents presented during the phase I which are the Software Requirement specification Software Quality Assurance plan along with the changes will be documented An updated time log will be documented The vision document will be updated to ensure that the requirements specification have captured the driving requirements of the project The updated version will be approved by the major advisor Cost Estimation The updated cost estimation will present the evaluated cost size and effort for the statistical analysis tool The revised version will be approved by the major advisor Project plan The project plan will be updated taking into consideration any new milestones corrected phases or iterations and deliverables along with the estimated completion date Implementation Plan The Implementation plan will define the activities and actions that must be accomplished during implementation The plan will include a 33 Work Breakdown Structure with time and cost estimates and the completion criteria Architecture design The complete design of the project wil
10. public java lang String setTitle java io PrintWriter out java lang String title 61 This method prints the title for the results page for all the tests Parameters out PrintWriter object title The title for the page setFooter public void setFooter java io PrintWriter out This method prints the footer for the results page for all the tests Parameters out PrintWriter object embedApplet public java lang String embedApplet java lang String appletName java lang String paraml java lang String param2 java lang String type This method returns a string containing the EMBED tag for the applet the Java version is set to 1 4 Parameters appletName the applet name to display param the X axis value gpa param2 the Y axis value salary 62 type the type correlation regression or scatterplot destroy public void destroy The Servlet Container will call this method when this servlet is being taken out of service Class RegistrationBean public abstract class RegistrationBean RegistrationBean is an abstract base class for accessing the database The RegistrationBean allows through home interface access to the database and the allows manipulation of the data 63 RegistrationRemote String moons String jopa double Salary double passwd String EDegree Boolean citizen Boolean 3 5 Boolean RoetAllUsers void BSigetLoginld
11. String BSigetGpa Double BSigetSalary Double BSigetPassword String BSigetDegree Char BSigetCitizenship Boolean BSigetJoinIn3Mths Boolean BSFindAIl Collection LinearRegression Bean Hypothesis Bean lt Correlation Bean BSFindUser String LiD String Pwd Collection y RegistrationHome EUN RegistrationBean Chi Square Bean Figure 12 Registration The detailed method description for this class 18 as follows getLoginId public abstract java lang String getLoginId getName public abstract java lang String getName getPassword public abstract java lang String getPassword 64 getSalary public abstract double getSalary getGpa public abstract double getGpa getDegree public abstract char getDegree getCitizen public abstract boolean getCitizen getJoinIn3Mths public abstract boolean getJoinIn3Mths setLoginId public abstract void setLoginId java lang String LId 65 setPassword public abstract void setPassword java lang String Passwd setName public abstract void setName java lang String name setSalary public abstract void setSalary double salary setGpa public abstract void setGpa double gpa setDegree public abstract void setDegree char degree setCitizen public abstract void setCitizen boolean citizen
12. BSiestimateY double X double BSihaveData int BSimembers double Figure 14 Correlation The detailed method description for this class is as follows getAllUsers public void getAllUsers throws java rmi RemoteException javax ejb CreateException 9 This method connects to the database and gets a collection of all the members in the database from the Registration bean getSalaryAndGpa public void getSalaryAndGpa throws java rmi RemoteException javax ejb CreateException This method gets the salaries GPA of the Masters and PhD students separately for analysis started public void started throws java rmi RemoteException javax ejb CreateException This method gets all the users from the database and starts the simple linear regression analysis for the Masters and Phd members addXY public void addXY double x double y throws java rmi RemoteException javax ejb CreateException 92 This method adds the x coordinates Parameters x The GPA of the member y The salary of the member Calculate public void Calculate throws java rmi RemoteException javax ejb CreateException This method calculates the slope y intercept coefficient of determination Correlation coefficient mean square due to error fvalue and the standard error for the Masters and PhD members separately getA public double getA throws java rmi RemoteException Returns the y inter
13. Figure 15 Chi Sguare 100 Figure 16 Table of Test data 121 Figure 17 Table of Test cases 123 Figure 18 Table of Hypothesis result 126 Figure 19 Table of Regression Result 128 Figure 20 Table 2 of Regression Results 127 Figure 21 Table of Chi Square Results 129 Figure 22 Phase I duration 151 Figure 24 Phase II duration 152 Figure 25 Phase III duration 152 CHAPTER 1 VISION DOCUMENT Introduction In today s world scientists and research professionals besides people from all wakes of life generate huge amounts of data in lots of different formats for different purposes The resulting volumes of data can no longer be analyzed by conventional technigues hence the need for statistical analysis tools The statistical analysis tool developed aims at providing graphical and tabular results to research guestions by performing the tests relevant to the guestions
14. Graph Result Filename Browse Log Errors Only j WorkBench Graphs to Display Data v Average vi Median v Deviation v Throughput 1593 ms 0 ms No of Samples 100 Latest Sample 406 Average 709 Deviation 248 Throughput 755 9531 minute fledian 656 For 30 users A Apache JMeter File Edit Run Options Help 2 Test Plan E questionsthreadgroup Graph Results HTTP Request Defautts Name lGraph Results ff HTTP Request Write All Data to a File Graph Results WorkBench Filename Browse 3 Log Errors Only Graphs to Display v Data v Average v Median v Deviation vi Throughput 3375 ms 0 ms a me No of Samples 300 Latest Sample 266 Average 1619 Deviation 559 Throughput 981 24725 minute Median 1375 128 For 45 users Apache JMeter File Edit Run Options Help 9 amp Test Plan questionsthreadgroup Graph Results HTTP Request Defaults Graph Results HTTP Request Write all Data to a File 8 s sa Filename Browse Log Errors Only Graphs to Display v Data v Average v Median vi Deviation vi Throughput No of Samples 450 Latest Sample 562 Average 2998 i Deviation 1542 Throughput 824 05005 minute Median 2579 The above given screen shot i
15. double S amp totalnoCitlnt int Cogpal double amp ssTotal double SchiSquareValue double memberSize int amp pcollection Collection GonoExpCitizen int alat sum double double POSEE int 0 1 3 1 sumXsquared double FValue double yesExplnt in RegistrationBean EosumY squared double amp SalaryPhD double amp bnoExplnt int W E a m br SgetUsers ad hBi BbcoefD double p it E msE msR adlie SstanChiSquarTestQ GbcoefC double amp sizeMS sizePhD int getO bservedValues StdEror double amp home RegistrationHome P aie end erenc es int obs int exp SgetA double started void SgetYesObsCitizen SaetB double SgetA double SgetNoObsCitiz en SgetSalaryAndGpa void 0 1 SgetB double SgetDegrees SgetAllUsers void Saddx Y double x double y void SgetYesObslnt started void SgetAllUsers void SaetNoObslnt SaddXY double x double y void SgetCoefDeterm double SgetTotalCitizen Calculate void SgetCoefCorel double SgetTotalint OgetCoeffCorrel double SgetStdErorEst double SaetTotalYesCitlnt SgetStdEmorEst double SestimateY double x double SgetTotalNoCitint SgetSalaryArray double haveData integer SgetYesExpCitizen getGpaArray double Sitems integer SgetNoExpCitizen SestimateY double x double SstatMSAnalysis wid SgetNoExplnt ShaveData int SgetPhDGpaAmay
16. role hypMember Hypothesis 0 1 role hypPerform end association chi between Registration 0 role chiMember Chisquare 0 1 role chiPerform end association cor between Registration 0 role corMember Correlation 0 1 role corPerform end CONSTRAINTS constraints Each member login id has to be unique context Registration inv uniquemember Registration alllnstances gt forAll ml m2 ml lt gt m2 implies ml username lt gt m2 username The GPA has to be between 0 and 4 45 inv GpaCheck Registration allInstances forAll m m gpa gt 0 and m gpa lt 4 The students should be either masters or PhD inv mastersOrPhd Registration alllnstances gt forAll m m degree Masters or m degree Phd The salary of an alumnicannot be null inv salaryCheck Registration alllnstances gt forAll m m salary gt 0 CORRELATION CLASS context Correlation inv atleastonememberforCorr Correlation allInstances forAll c c corMember gt size gt 1 inv xynotemptyforCorr Correlation allInstances gt forAll c c corMember gpa gt size gt 1 and c corMember salary gt size gt 1 REGRESSION CLASS context Regression inv atleastonememberforRegress Regression allInstances forAll r r testMember gt size gt 1 inv xynotemptyforRegress Regression allInstances forAll r r testMember gpa gt size gt 1 and r testMember salary gt size gt 1 HYPOTHESIS CLASS context Hyp
17. the probability of getting a job being dependent on his her citizenship and the dependence of salaries on the degrees obtained The object oriented design methodology and programming technigues are applied to the development of the software tool The project applies the fundamentals of the J2EE architecture and the Enterprise Java Beans construct towards building a web based tool which provides the registered user with the ability to perform the aforementioned tests on the alumni data The main features of the project are an input data set of the KSU statistics alumni in Oracle database which provides the reguired information for the statistical analysis tests Internet access and easy to use interface to use the application and instant graphical representation and tabular summarization of the analysis results TABLE OF CONTENTS TABLEOPCONTENTS ssa a Kaa ee CHAPTER 1 VISION DOCUMENT sities ttes terne CHAPTERS 2 PROJECT PLAN ua dta b c pH 4 CHAPTER ARCHITECTURE DBSIQGN itin I neue 36 CHAPTER 4 COMPONENT 22 7 1 50 CHAPTER 5 SOFTWARE QUALITY ASSURANCE 107 CHAPTER 6 TEST PLAN vsem tamas assa Rd teal Eo CHAPTER 7 ASSESSMENT EVALUATION 118 USERS MANUAL LAO CHAPTER 9 PROJECT EVALUATION se JAT CHAPTER 10 FORMAL TECHNI
18. Overview of the system The objective of the project is to develop a website for registered members to perform different statistical analysis on the member data The tests being conducted are the regression analysis the chi square test the correlation test and the hypothesis test The system is so designed that each test is an enterprise java bean Thus there are four session beans namely LinearRegressionBean HypothesisBean CorrelationBean ChisquareBean There is one entity bean RegistrationBean that creates a table in the database called RegistrationBeantable on deployment of the system The system has one servlet and uses java applets to output graphs The functions of each of these are explained below Class RunRegressionServlet RunRegressionServlet is a servlet which creates instances of all the session beans and the entity bean whenever the user selects a research question The servlet basically facilitates e User Registration e User Login e Linear Regression Calculation e Correlation Calculation e Chi Square Calculation Hypothesis Calculation 50 The servlet also includes some helper functions for displaying the output to the web Servet BSicreateRegression double BSiCreateCorrelation double BSiCreateHypothesis double BSiCreateChiSquare double BSiRegisterUser double PLI N RegistrationBean ChisguareBean
19. Two Sample for Means Masters PhD Mean 61043 5 84077 0 Variance 1 633144339E8 3 131137133E7 Observations 6 4 Hypothesized Mean 0 Difference df 8 t Stat 3 3448 Pooled SD 10668 331 e Correlation vids Department Of Statistics Es Alumni Center Graph Mode Ots Mem salary 80 000 60 000 ZEN R Sguare 0 734 117025 842 52167 895 Equation 117025 842 52167 895 11103 286 R Square 0734 P value 0 001545 e Chi Square test 145 sir Department Of Statistics NEN bo Tome JA EN Alumni Center Results Chi Sguare Test of Independence 1 Citizens Internationals Total Join in 3 Months Yes 3 2 5 observed Join in 3 Months Yes 2 2 NA expected Join in 3 Months No 2 3 5 observed Join in 3 Months No 2 2 NA expected Total 5 5 10 Degrees of freedom 1 Chi Square 0 0 Critical Value CV for 5 level of significance 3 841 If Chi Square value is greater than CV than reject the null hypothesis If not than we fail to reject the null hypothesis System Configuration Oracle Database Server is used for storing all the persistence data such as the data pertaining to the members that they enter into the system when they register Sun ONE Application Server 7 0 is used as J2EE container which in
20. Validity and integrity of each of the statistical analysis The calculations that are needed are to be validated for performance of the right kind of analysis and in the right manner The live data from the alumni center may not be possibly available There are some present issues with the alumni foundation providing the data to the project The learning curve The project involves use of many new technologies and the need to understand and use them to meet the reguirements Mitigation Approach The research will be done in the early part of the project to get all the information about the statistical analysis methods that need to be present on the website This will help in identifying the correct solution in the early part of the project The data if not available test data will be used to demonstrate and test the project The technologies will understood completely and their need identified during the requirements specification phase and also during the design phase so that the learning curve is not too steep Constraints The project mainly deals with the development of the website using J2EE for the application server and EJB to develop the business logic needed for the project Thus the main constraints will deal with the learning curve to understand the methodologies and then developing the application in the same the features need to be implemented well The implementation strategy will emphas
21. assist in giving a head start to developers who are new to this technology in the CIS department of Kansas State University Installation Guide 1 Download and install the J2SE SDK 1 3 1 from http java sun com j2se 1 3 and J2EE SDK from http java sun com j2ee download htmlZsdk If you need to use build tool then download Apache ant from http ant apache org bindownload cgi 1 Set the environment variables to the values noted in the following table Table Setting for Environment Variables Environment Variable lt JAVA HOME gt The location of the J2SE SDK installation lt J2EE HOME gt The location of the J2EE SDK installation 130 lt _ gt The location of the ant installation PATH Should include the bin directories of the JOEE SDK J2SE SDK and ant 2 If you use Oracle Database Server then you will have to copy classes111 zip from C Oracle ora90 jdbc lib into lt JAVA HOME gt lib system You can also download classes111 zip from the Oracle website at http otn oracle com software tech java sglj jdbc index html Starting The J2EE Server The command to start the J2EE Server is j2ee verbose Starting The Deploytool tool For the GUI version of the deploytool the command to start the tool is deploytool after the j2ee startup is complete Creating The J2EE Application The application generally comprises of three components an enterprise bean a J2EE application client and a Web compo
22. at the end of each phase to the major professor and the committee for review Committee The responsibilities of the committee include Timely review of the student s work and progress in the project Giving feedback and suggestions to guide the developer Documentation The essential documents reguired are described below Minimum documentation reguirements Software reguirements specification 109 The SRS lists the reguirements of a system and it should correctly describe the system reguirements It specifies the functionality and performance of the project in terms of speed reliability etc It describes how the system interacts with the people using and also specifies the design constraints and standards to be followed Software verification amp validation plan The purpose of this document will be to develop and record formal procedures for the verification and validation of the statistical analysis tool Formal Specification Document A section of the tool s design will be formally specified using Alloy or OCL Software architecture design document This document describes the overall structure of the statistical analysis tool An object model will be constructed using rational rose which will describe the various classes used in the project It will also consist of an interaction diagram depicting the key interactions of the software Engineering Notebook The engineering book consists of Gantt chart of the
23. components The J2EE specification defines the following J2EE components e Application clients html code and applets are components that run on the client Java Servlet and Java Server Pages JSP technology components are Web components that run on the server e Enterprise JavaBeans EJB components enterprise beans are business components that run on the server The Java application servers are based on the Java 2 Platform Enterprise Edition J2EE This platform allows the users to completely focus on the implementation of the business logic and not worry about the low level details which are managed by 2 The J2EE server communications are summarized in the figure below Client Tier Web Browser Web pages applets and optiona Web Tier JavaBeans class Business Tier Application Client and optiona JavaBeans class 2 Server Figure 1 J2EE Server The client can communicate with the business tier either directly or through the web tier In the statistical analysis tool the client will communicate through the web tier The web tier constitutes of either servlets or Java Server Pages The business tier constitutes of the business logic implementation The Enterprise Java Beans which comprise the business tier handle this responsibility Since it is based on the Java framework the J2EE architecture is platform independent and component based which makes it easy to
24. enough members to perform the test Message There are no MS members to perform the test Message There are no PhD members to perform the test Chi Sguare Bean No Citizens No International students No person with a job in 3 months of graduation No person without a job 113 months of graduation Message There are no members currently in the database who are citizens Please try again later to perform the Chi Sguare test Message There are no members currently in the database who are citizens Please try again later to perform the Chi Sguare test Message There are no members currently in the database who got a job in 3 months of graduation Please try again later to perform the Chi Square test Message There are no members currently in the database who did not get a job in 3 months of graduation Please try again later to perform the Chi Square test Hypothesis Bean No MS alumni There are no MS members currently in the database to perform the Hypothesis test Please try again later No PhD alumni There are no PhD members currently in the database to perform the Hypothesis test Please try again later Correlation No members There are no members currently in the database to per form this test Figure 17 Table of Test cases Since the tool performs statistical analysis the results obtained using it were compared with the results obtained using Microsoft Excel
25. for anticipated future changes The Inspectors gave a partial rating for defect conditions messages being specified completely but said that all necessary defect conditions messages have been specified Even so a paragraph was added in the Analysis requirement section 3 3 3 of the SRS to incorporate all error messages The letters submitted by the inspectors to confirm their involvement in the formal technical inspection will be documented and submitted with the rest of the documentation 158 REFERENCES IEEE Std 1028 1998 Standard for Software Reviews and Audits 1998 Edition IEEE 1983 IEEE Std 730 1998 IEEE Standard for Software Quality Assurance Plans 1998 Edition IEEE 1998 IEEE STD 830 1998 IEEE Recommended Practice for Software Requirements Specifications 1998 Edition IEEE 1998 Pressman Roger S Software Engineering A Practitioner s Approach Fifth Edition Mc GrawHill NY June 2001 Developing Enterprise Applications Using the J2EE Platform java sun com 2003 IEEE Std 829 1983 IEEE Standard for Software Test Documentation 1983 Edition IEEE 1983 Java Applets for graph West Virginia University stats web site http www stat wvu edu SRS Modules Applets Dr Scott Deloach s CIS748 lecture notes http www cis ksu edu sdeloach 748 IEEE guide for software quality assurance planning 730 1 1995 159 Lecture notes CIS 748 Software Management Dr
26. g into the design code and test documentation EE Is each requirement uniquely and correctly identified Understandability Questions Yes No Partial Is each requirement written in clear Yes concise unambiguous language 156 Does the SRS contain only necessary Yes implementation details and not contain unnecessary details Is it over specified Maintainability Does the documentation follow MSE Yes Is the documentation clear Yes Verifiability Testability Questions Yes No Partial Are the requirements verifiable 1 e can Yes Can be verified the software package be checked to see whether requirements have been fulfilled Clarity Questions Yes No Partial Are all of the decisions dependencies and assumptions for this design documented one clear meaning Is each statement written as clearly as Y possible Functionality Yes Is each concept defined only once with Yes es 157 Ouestions Yes No Partial Does the design implement the Yes specifications and requirements Reliability Are the defect conditions codes messages Partial Necessary specified completely and meaningfully defect conditions messages are specified Evaluation The inspection resulted in minor changes in the SRS The SRS did not define the requirements for which future changes were to be anticipated This was amended in the version 2 0 of the SRS by adding a section 2 6
27. of the member degree The Masters or PhD degree obtained by the member citizen The citizenship of the member salary The salary of the member join The variable indicating if the member joined a job within 3 months of graduation findUser public boolean findUser java lang String loginId java lang String password throws java rmi RemoteException javax ejb CreateException This method looks up a member in the database given his loginId and password Parameters Loginld The username with which the member logs in to do the analysis 71 Password The password used by the member to log in getAllUsers public void getAllUsers throws java rmi RemoteException javax ejb CreateException This method gets a collection of all the members in the database from the Registration bean getSalaryAndGpa public void getSalaryAndGpa throws java rmi RemoteException javax ejb CreateException This method gets the salaries and GPA of the Masters and PhD students separately for analysis started public void started throws java rmi RemoteException javax ejb CreateException This method gets all the users from the database and starts the simple linear regression analysis for the Masters and Phd members 72 startMSAnalysis public void startMSAnalysis throws java rmi RemoteException javax ejb CreateException Start analysis for MS members startPHDAnalysis public void startPHDAnalysis throws java rmi RemoteExcept
28. or the PhD students 69 RegressionRemote BSigetA double BSigetB double BSladdXY double x double void BSigetAllUsers void BSigetCoefDeterm double Servet BSigetCoefCorrel double BSistarted void BSigetStdErrorEst double RestimateY double x double 7 BSihaveData integer BSiitems integer BSistartMSAnalysis void RegistrationBean BSigetPhDGpaArray double BSistartPhDAnalysis void BSigetPhDSalaryArray double BSigetMSError double BSigetSSError double BSigetSalaryAndGpa void BSigetFValue double BSigetSSTotal double BSigetSSRegress double BSigetMSRegress double BSigetMSSalaryArray double BSigetMSGpaArray double RitemsMS int RitemsPhD int BSicalculate void Figure 13 Regression The detailed method description for this class is as follows addUser public boolean addUser java lang String Loginld java lang String Passwd java lang String name double salary double gpa char degree boolean citizen boolean join 70 throws java rmi RemoteFxception javax ejb CreateException This method adds a new member to the database Parameters LoginId The username with which the member logs in to do the analysis Passwd The password used by the member to log in name The member s name salary The salary of the member gpa The GPA
29. output of each test separately But again since the user can select only one test at a time the output can be considered to be a single output The weighting factor for selecting a research question and Member registration as input was changed from average to simple On the other hand the weighting factor of Graph Table of analysis test as output was change from simple to average The changes are reflected in the table below 29 Member Login nct Member Registration Select a research question Outputs Member login confirmation Member Registration 4 confirmation Graph Table of analysis test i Dad Fies Linearregression ime IChsmaews a E User to application server o Figure 7 Table of Unadjusted function points After the changes were made the new total unadjusted function points UFP came up to be 90 Calculate Adjusted Function Point To calculate the Complexity adjustment value several factors have to be considered such as Backup and recovery code design for reuse etc All the factors and their estimated values in this project are shown in the following table The adjusted function point denoted by FP is given by the formula FP total UFP 0 65 0 01 Total complexity adjustment value or 30 total Complexity adjustment factor Total complexity adjustment value is counted based on responses to guestio
30. setHypothesisTable public void setHypothesisTable java io PrintWriter out double meanMS double meanPHD double varianceMS double variancePHD int nMS int nPHD int degrees double tValue double CDF 59 This method creates the Hypothesis XML tree and then uses XSL to display it to the web Parameters out PrintWriter object meanMS means of salries of all MS members meanPHD means of salaries of all PhD members varianceMS variance for MS variancePHD variance for PHD nMS number of MS nPHD number of PHD degrees Degrees of freedom tValue the t Value CDF the CDF value setExtraRegressionTable public void setExtraRegressionTable java io PrintWriter out double correlation double determine double stderror int size double ssError double ssRegress double msError 60 double msRegress double ssTotal double slope double intercept This method creates the Regression XML tree and then uses XSL to display it to the web Parameters out PrintWriter object correlation coefficient of correlation determine coefficient of determination stderror Standard Error size number of MS PHD members ssError Sum of Squares for error ssRegress Sum of Squares for regression msError Mean Squares for error msRegress Mean Squares for regression ssTotal Sum of Squares total slope slope for the best fit line intercept intercept for the best fit line setTitle
31. the testing will be suspended until the bug is fixed 8 2 Resumption Criteria Testing will be done again to ensure that the bug has been fixed and the selected tests give the expected results Only then will the testing be resumed 9 Deliverables The following artifacts will be delivered in the testing process 1 Test plan 2 Test cases 117 3 Test log 10 Responsibilities Since there is only one person developing the SAT the developer is responsible for all the testing activities to be carried out 118 CHAPTER 7 ASSESSMENT EVALUATION Testing In order to test whether all the analysis tests gives the right results a table of know values was inputted into the database Now the known results were compared with the expected results from the statistical analysis tool for validity The results were also checked with the excel output for the same data The project was also tested for performance using the tool JMeter JMeter is a subproject of Jakarta and is a Java based tool for load testing client server applications The results obtained from this tool are also documented in this document Given below is the table of known data that was entered into the database Padmaja pad 6 156432 35 Yes No Yes sad 3 6 No Yes No es No sab Figure 16 Table of Test data Test cases The tool requires members to register before they can do the analysis There is a membership form pro
32. to a J2EE application you must redeploy the application Stopping the J2EE server The command to stop the server is J2ee stop 139 CHAPTER 8 USER S MANUAL General Information System Overview The Statistical Analysis Tool is a web based application for registered users to perform different statistical analysis on the user data The website will allow four types of analysis to be done on the data the regression analysis the chi square test the correlation test and the hypothesis test The website in particular will be able to provide analysis for questions of the nature of e Is there arelationship between the grade point average GPA of an alumni member and the salary bracket that he currently belongs to e Is there a difference between the average salaries of the Master s student and the average salaries of the PhD students at a 5 level of significance e Does higher GPA result in higher salaries for masters students e Does higher GPA result in higher salaries for PhD students e Is getting a job within 3 months of graduation independent of the citizenship of the student at the Masters level Using the System The user has to enter the URL http procyon cis ksu edu 8000 lastapp indexproj htm to take him to the system In order to use the system and perform analysis the user has to log in to the system This can be done only if the user has registered to be a member of 140 this website An online member
33. which is known to give reliable and accurate results The same data set shown in Table 17 was used in both cases The results obtained for the four tests are summarized below 1 Hypothesis test Known values Excel output t Test Two Sample Assuming Variances Equal Variable 1 Variable 2 121 Variance Observations Pooled Variance Hypothesized Mean Difference df t Stat Expected values using tool 61043 5 84077 163331531 9 31302572 6 4 113820671 9 0 8 3 344683159 t Test of Independence for Two Sample Means Masters PhD Mean 61043 5 84077 0 Variance 1 633144339E8 3 131137133E7 Observations 6 4 Hypothesized Mean 0 Difference df 8 t Stat 3 3448 P one tail 0 0051 Figure 18 Table of Hypothesis result Result of test The tool was successful in computing the test 2 Correlation The result for this test is evident in the output graph itself The graph shows the points with the members GPA as the independent variable and the Salary as the corresponding dependent variable The results were also checked with excel Known Values Excel output Correlation Co efficient R square 0 85675 0 73421 122 Observed Values using tool Correlation Co efficient 0 857 R sguare 0 734 Result of test The tool was successful in computing the test 3 Simple Linear Regression Simple linear regression analysis is performed on the Masters and PhD data separatel
34. 0 students henceforth If we include the students already graduated and the current students 21 faculty besides other users using the website the number of users will be approximately 500 Oracle is scalable enough to meet these reguirements for the next 20 years The number of connections to the system should not slow down the application to a large degree The data for the analysis will be obtained from the database of users so the response time for a guery from the client side to the database side should not be more than 5 seconds Error handling should be implemented and the application should be able to handle all run time errors The application should be flexible for future enhancements for example the addition of a few more research analysis questions Quality attributes The application will be composed of Enterprise Java Beans which serve as the business logic layer and Java servlets and JSP pages which serve as the web tier the code will be documented using current standards and documentation will be provide to the user to make use of the application Added documentation will include the System Delivery Document which will document the installation procedure Remote availability will depend on an active Internet connection Testing for the application will be followed as a continuous process followed in parallel to the requirements design and implementation phase This will include a pre t
35. 00 This method gets the observed number of citizens among the alumni Returns Number of observed citizens getNoObsCitizen public int getNoObsCitizen throws java rmi RemoteException This method gets the observed number of non citizens among the alumni Returns Number of observed non citizens getDegrees public int getDegrees throws java rmi RemoteException This method gets the degrees of freedom Returns 1 as the degrees of freedom for a 2 x 2 table is 1 getYesObsInt public int getYesObsInt throws java rmi RemoteException 101 This method gets the observed number of alumni who have got jobs within three months of graduation Returns Number of alumni with jobs within 3 months of graduation getNoObsInt public int getNoObsInt throws java rmi RemoteException This method gets the observed number of alumni who did not get jobs within three months of graduation Returns Number of alumni without jobs within 3 months of graduation getT otalCitizen public int getTotalCitizen throws java rmi RemoteException This method gets the total number of citizens among the alumni Returns Number of citizens getTotalInt public int getTotalInt throws java rmi RemoteException 102 This method gets the total number of alumni Returns number of alumni getTotalYesCitInt public int getTotalYesCitInt throws java rmi RemoteException This method gets the total number of citizens who got jobs within
36. 24 Regression Statistics Multiple R 0 962026626 R Sguare 0 92549523 Observations 4 ANOVA df SS MS Regression 1 86911143 2 86911143 2 Residual 2 6996572 8 3498286 4 Total 3 93907716 Coefficients Standard Error Intercept 76437 2 32217 10066 X Variable 1 41692 8364 551871 Observed Values using tool SUMMARY OUTPUT Regression Statistics Multiple R 0 962 R Square 0 9255 Standard Error 8364 5519 Observations 4 Intercept 76437 2 Slope 41692 0 ANOVA df SS MS Regression 1 8 69111432E7 8 69111432E7 Residual 2 6996572 8 3498286 4 Total 3 9 3907716E7 125 Figure 20 Table 2 of Regression Results Result of test The tool was successful in computing the test 4 Chi Square Test Expected Values The test was manually computed and the results were Ho Citizenship and getting a job in three months of graduation are independent Ha Citizenship and getting a job in three months of graduation are not independent Chi Square test value Sum E 2 E O Observed value E Expected Value 3 2 5 2 2 5 2 2 5 2 2 5 3 2 5 2 2 5 2 2 5 2 2 5 0 4 E row total Column total Sample Size Critical Chi Square value 3 841 We accept the null hypothesis Observed Values Chi Square Test of Independence Citizens Internationals Total Join in 3 Months Yes 3 2 5 observed Join in 3 Months Yes 2 5 2 5 expected Join in 3 Months No 2 3 5 observed Join in 3 Months No 2 5 2 5
37. 748 SoftwareManagement CIS 771 Software Specification The developer also has sufficient knowledge of Statistics 114 CHAPTER 6 TEST PLAN 1 Test Plan Identifier MSE Version 1 0 2 Introduction This document will describe the test plan for the Statistical Analysis Tool which is being developed in partial fulfillment of the requirements of the MSE degree The purpose of the document is to draft a test approach for testing the critical aspects of the statistical analysis tool and the overall strategy to be adopted in the process The vision document and the Software Quality Assurance plan will be used as a reference 3 Test Items The items to be tested for the Statistical Analysis Tool SAT are 1 Statistical Analysis Tool Application 2 The critical requirements specified in the Software Requirements Specification Sections 3 3 3 4 and 3 5 4 Features to be tested The following features of the SAT will be tested 1 The login Registration and test analysis features which are described in detail in Sections 3 3 1 3 3 2 and 3 3 3 of the SRS Also the features described in sections 3 4 and 3 5 2 The Statistical Analysis Tool program structure 115 3 The functionalities of the session beans and the entity bean which perform the tests and the registration 5 Features not to be tested Not Applicable as all features will be tested 6 Test Cases 6 1 Login e f the member enters incorrect input like empty
38. CAL INSPECTION 154 LIST OF FIGURES Figure 1 J2EE 7 Figure 2 Overview of the Statistical Analysis Tool SAT 10 Figure 3 Use Case of Member Functions 20 Figure 4 Use Case of system administrator Functions 20 Figure 5 Gantt chart 23 Figure 6 Table of Function types 28 Figure 7 Table of Unadjusted function points 29 Figure 8 Table of complexity adjustment value 30 Figure 9 Table of COCOMO formulae 3l Figure 9 Object Model 37 Figure 10 Servlet interaction 53 Figure 11 Registration 66 Figure 12 Regression 72 Figure 13 Hypothesis 84 Figure 14 Correlation 93
39. Client Jar File Edit Tools Help KYS ejela miejala Files Applications Newstat S lejielsse Please select the object to be deployed and the server to which it should be deployed Object to Deploy Newstats i amp servers Tari get Server o localho CANTE The server can send back a client JAR file It contains the extra RMI IIOP stub classes that client applications written to access this application will need at runtime 1 Return Client Jar Client JAR File Name Ihome grads padmaja 23Nov03 javaZip NewstatsClient jar Browse sack Next gt Finish Save object before deploying e Click Next e In the JNDI Names dialog box verify the names you entered in the previous section e Click Next e In the WAR Context Root dialog box enter the Context Root field When you run the Web client the context root will be part of the URL e Click Next e In the Review dialog box click Finish 138 e In the Deployment Progress dialog box click OK when the deployment completes Running The Web Client To run the Web client type the following URL in the browser with all relative information http lt localhost gt 8000 lt context root gt Modifying The J2EE Application Since the J2EE SDK is intended for experimentation it supports iterative development Whenever you make a change
40. PhD students 3 Hypothesis testing Test to see if there is a difference in the average salaries of the Master s student and the PhD students at a 5 level of significance 4 Chi Square test Is getting a job within 3 months of graduation independent of the citizenship of the student at the Masters level model StatisticalAnalysis 39 classes The servlet class creates instances of the Registration bean Regression bean Correlation bean hypothesis bean and the chi sguare bean when the user reguests the particular test to be performed or registers class servlet operations createRegression r Regression createCorrelation co Correlation createHypothesis h Hypothesis createChiSguare c Chisguare RegisterUser u Registration end The Registration class is an entity bean which creates the member table in the database When the member registers the information is sent by the servlet to the Registration bean which adds the record to the database class Registration attributes name String username String password String salary Integer gpa Real degree String citizen Boolean joinin3mths Boolean end 40 class AllRegistrations operations addMember m Registration DeleteMember m Registration SearchMember id String Set Registration if memberEntity gt exists m Registration m username id then memberEntity gt select m Registration m u
41. Project Plan Cost Estimation using the COCOMO model and the function point analysis Time Log 110 User Documentation The user documentation will consist of User manual which will identify all error messages and program limitations along with their corrective measures Installation guide Source code and testing code along with any other files used The following documents will be provided at the end of each phase by the Developer Phase I Software Requirement Project Overview Software Requirements Specification Gantt chart of Project plan Cost Estimate Prototype of the tool Software quality assurance plan Phase II Software Design Formal Requirement Specification Architecture design 111 Test plan Implementation plan Phase III Software Implementation User Manual Testing Evaluation Project Evaluation References Source Code Standards practices conventions amp metrics The MSE project portfolio will serve as a guideline in developing the documents Documentation Standards The IEEE software engineering standards were observed in designing the software requirements specification SRS and the software quality assurance plan SQAP Coding Standards The source code will follow the guidelines in the Java coding standards Metrics 112 Function point analysis and Source lines of code will be a measure of the size of the tool The COCOMO model will be a measure of the programmer mont
42. STATISTICAL ANALYSIS TOOL USING ENTERPRISE JAVA BEANS AND THE J2EE ARCHITECTURE by PADMAJA HAVALDAR B Sc University of Goa 1995 M Sc University of Goa 1998 A REPORT Submitted in partial fulfillment of the requirements for the degree MASTER OF SOFTWARE ENGINEERING Department of Computing and Information Sciences College of Engineering Kansas State University Manhattan Kansas 2003 Approved by Major Professor Dr Daniel Andresen ABSTRACT Statistical Analysis has been used to provide solutions to complex problems in such diverse areas as communications stability finding patterns in data sets of variables and other areas of interest Although these problems previously lacked adeguate mathematical treatment the results of statistical analysis have been significant both for explanation of a pattern and also for prediction Statistical analysis is a general scientific method and it has applications in many areas of scientific research The statistical analysis tool developed provides such information about the department of statistics alumni at Kansas State University This project consists of implementing a statistical analysis tool which provides patterns to explain research guestions pertaining to the KSU statistics alumni The project uses statistical analysis methods like linear regression correlation hypothesis and chi sguare tests to provide the relationship between GPA of the alumni his her degree MS PhD
43. Scott Deloach Spring 2002 Formal Technical Checklist http www csc calpoly edu jdalbey 205 Resources InspectChecklist html Formal Technical Checklist http satc gsfc nasa gov fi IEEE 1042 Guide to Software Configuration Management Plans 160
44. a website which will enable users to register and perform different statistical analysis on the member data The tool will essentially perform four kinds of analysis namely regression correlation hypothesis test and chi sguare test depending on 11 the research guestions specified on the alumni website The output will then be displayed in the form of tables or graphs on the webpage Goal The website in particular will be able to provide analysis for guestions of the nature of 1 Correlation Is there a relationship between the grade point average GPA of an alumni member and the salary bracket that he currently belongs to 2 Hypothesis testing Is there a difference between the average salaries of the Master s student and the average salaries of the PhD students at a 5 level of significance 3 Linear Regression Does higher GPA result in higher salaries for masters students Does higher GPA result in higher salaries for PhD students 4 Chi Square test Is getting a job within 3 months of graduation independent of the citizenship of the student at the Masters level Technology The application uses a three tier application for the development of the web site The web site uses the J2EE server for deploying the Jar files which contain the EJB s that provide the business logic The first tier or the client tier includes dynamic html web pages and applets that run on the client The web tier consists of the servlet for t
45. ang while importing ear files and would crash the complete system when closed Alumni data The Analysis tool had to perform analysis on the alumni of the department of statistics This data is available at the Kansas state university alumni center But the developer even after much trying could not get access to this data due to security reasons So for the purpose of the MSE project dummy data is used instead of the actual alumni data 148 Number of lines of code In the First phase of development the estimate for the lines of code was 4636 This estimate was made based on the function point analysis It should be taken into account that the developer had no prior experience besides an academic assignment in developing such a project Thus the estimate for the number of lines of code differed in the second phase They were estimated to be 3659 This was de to the fact that in the first phase the developer had considered the selection of research guestion for each test as an input But essentially the user can choose only one research guestion at a time hence the selection of a research guestion is considered as one single input This reduced the lines of code estimate to 3659 The actual number of lines of code came out to be 2800 The tool makes use of graphs to output the results The developer based these graphs on a source which was available on the internet and made re use of this code This reduced the number of lines of code whi
46. cept Returns Returns the y intercept getB public double getB 93 throws java rmi RemoteException Returns the slope Returns Returns the slope getCoefDeterm public double getCoefDeterm throws java rmi RemoteException Returns the coefficient of determination Returns Returns the coefficient of determination getCoefCorrel public double getCoefCorrel throws java rmi RemoteException Returns the Correlation coefficient Returns Returns the Correlation coefficient getStdErrorEst public double getStdErrorEst throws java rmi RemoteException 94 Returns the standard error Returns Returns the standard error getSalaryArray public double getSalaryArray throws java rmi RemoteException Returns the Salary values array for all members Returns Returns the Salary values array for all members getGpaArray public double getGpaArray throws java rmi RemoteException Returns the GPA values array for all members Returns Returns the GPA values array for all members estimateY public double estimateY double x throws java rmi RemoteException This method calculates the equation of the best fit line 95 Parameters x The x coordinate i e the GPA Returns Returns a b x the eguation of line haveData public int haveData throws java rmi RemoteException javax ejb CreateException This method checks to see 1f we have enough data to do the test as we require at least 3 poi
47. ch the developer would then have to write The distribution of the lines of code is as follows Entity Java Beans 1869 Servlet 1040 XSL 120 Total 3029 lines of code Cost estimation 149 The cost estimation was done using the function point analysis and COCOMO model Since the project plan and cost estimate was refined in the second phase the programmer months also changed from the earlier estimate of 11 person month to 9 person month The actual time reguired for the completion of the project was 9 25 months which is fairly consistent with the estimate from the second phase The first and second phases of the project took longer than the third phase This was due to the initial unfamiliarity with the technologies and the unexpected problems encountered while using the deploy tool in the first phase and due to the design in the second phase The project uses many mathematical formulae to develop the tests and it was not clear how this should be done using OCL and the USE tool The problem was then resolved after valuable input from the committee members Project duration The following table shows the duration of the project i 1 N 03 15 03 06 30 03 07 01 03 10 27 03 10 28 03 12 08 03 The breakdown of activities in each phase is summarized in the following table and pie charts 150 minutes minutes minutes si N meetings Deploying amp 1110 810 1800 Testing 40 Phase I Duration
48. cludes EJB Server and Web Server to maintain Enterprise JavaBeans and Web components For the Client the hardware requirement is at least Intel Pentium III 300 Mhz or 1 0 Ghz Athlon which has at least 256 MB RAM and has at least 200 MB freed hard disk space To run the application Netscape Navigator and Microsoft Internet Explorer versions that support Java Script and Java applets are required 146 CHAPTER 9 PROJECT EVALUATION Problems Encountered Learning curve The Statistical analysis tool analyses four kinds of tests namely the hypothesis test chi square test simple linear regression and correlation The developer had to study these tests in order to implement them Statisticians have come up with a table of probabilities for the t value p value and chi square values required to come up with the critical value which is compared to the test value to obtain the result The developer had to find a way of implementing these probabilities and had to do research on ways to do it The J2EE architecture also posed a challenge in the learning process since the developer had not worked extensively with this technology prior to this project In the earlier part of the project a lot of time was spent on understanding this and the various class paths to be set J2EE and Deploy tool Does not update files automatically The documentation for the deploy tool suggests that if we change any file within the EJB jars or the web component and
49. compile the changes we just need to click on update files to get the latest version for all files But the deploy tool does not do this on a regular basis The files are not updated The only way to be sure that the deploy tool gets the updated version of your files is to load them again into the WAR or JAR and redeploy them 147 Not best suited for unit testing or development practices If the development environment has considerations of space or memory the deploy tool is not best suited for unit testing and deploying many times with changes The problem is that even with few EJB jars and just one WAR component the EAR file becomes huge and it is backed up each time you redeploy The space crunch causes the deploy tool to perform erratically giving errors like Out of Space exception while deploying your application After some amount of research I found a clean up tool for J2EE which cleans up the repository directory in the J2EE server EJB packaging error occurs in deploy tool when Finish button is clicked too early When deploying even if all your properties are set and you can hit Finish to start deploying it is best to keep clicking Next until the last screen of the wizard is reached Some form of corruption takes place in the EAR file if this is not followed and the next time you start deploy tool it complains that the EAR file could not be loaded as it is corrupted There were other issues where the deploy tool used to h
50. d deviation getTValue public double getTValue throws java rmi RemoteException Returns the t Value Returns returns the t Value getCDF public double getCDF throws java rmi RemoteException Returns the CDF Returns returns the CDF 89 ejbCreate public void ejbCreate setSessionContext public void setSessionContext javax ejb SessionContext ctx ejbRemove public void ejbRemove ejbActivate public void ejbActivate ejbPassivate public void ejbPassivate ejbLoad public void ejbLoad ejbStore public void ejbStore 90 Class CorrelationBean public class CorrelationBean CorrelationBean is a session bean which handles the computation for the correlation test It tests to see if higher GPA results in higher salaries for all members CorrelationRemote E3home RegistrationHome coll Collection G8Salary double G8gpal double memberSize int Servet sumxX sumY double G3sumXsquared double G8sumYsquared double j sumXY double po b double N double double E5SstdError double RegistrationBean BSigetA double BSigetB double BSigetSalaryAndGpa void BSigetAllUsers void BSistarted void double x double void calculate void BSigetCoeffCorrel double BSigetStdErrorEst double BSigetSalaryArray double BSigetGpaArray double
51. dDouble public double roundDouble double d This method rounds double input to 4 decimal places Parameters d double value to be rounded Returns double value after rounding to 4 decimal places roundDouble2 public double roundDouble2 double d This method rounds double input to 2 decimal places 55 Parameters d double value to be rounded Returns double value after rounding to 2 decimal places calculateRegression public void calculateRegression java io PrintWriter out throws java io IOException This method get the result for linear regression calculation from the session bean and displays to the webpage Parameters out PrintWriter object Throws java 1o IOException the servlet is interrupted is IOException occurs calculateHypothesis public void caleulateHypothesis java i0 PrintWriter out throws java io IOException This method get the result for linear regression calculation from the session bean and displays to the webpage Parameters out PrintWriter object 56 Throws java io IOException the servlet is interrupted is IOException occurs calculateChiSquare public void caleulateChiSguare java io PrintWriter out throws java io IOException This method get the result for Chi Square calculation from the session bean and displays to the webpage Parameters out PrintWriter object Throws java io IOException the servlet is interrupted is IOException occurs setChiSquareT abl
52. double QgetYesExpiniQ members double OstatPhDAnalysis void getChiSquareValue yetPhDS alaryArray double SgetMSEror double QoetS5 Eroi double getSalaryAndGpa void F getF Valued double CorrelationHome S getSSTotal double getSSRegress double SgetMSRegress double getMSSalaryAmay double SgetMSGpaArray double ChiBean RegressionHome N RegressionBean Figure 10 Object Model 38 HypothesisHome N HypothesisBean Formal Reguirement Specification A model to test and formally specify primary properties of the SAT was developed Object constraint language OCL was used to write the constraints and the USE tool was used to test the model The model designed is given below The four tests are used as four classes namely the correlation class regression class chi sguare class and the hypothesis class The registration class also has an All Registration class which acts as a collection of data which is passed to the session beans The model was tested with command scripts to check if the constraints behaved as expected Research guestions 1 Correlation Does higher GPA result in higher salaries 2 Linear Regression Does higher GPA result in higher salaries for master s students Does higher result in higher salaries for
53. e public void setChiSquareTable java io PrintWriter out int yesObsCitizen int noObsCitizen int degrees int yesObsInt int noObsInt int totalCitizen int totalInt 57 int total Y esCitInt int totalNoCitInt int yesExpCitizen int noExpCitizen int noExpInt int yesExpInt double chiSquareValue This method creates the Chi Square XML tree and then uses XSL to display it to the web Parameters out PrintWriter object yesObsCitizen total number of citizens who joined a job within 3 months observed noObsCitizen total number of citizens who did not join a job within 3 months observed degrees Degrees of freedom yesObsInt total number of internationals who joined a job within 3 months observed noObsInt total number of internationals who did not join a job within 3 months observed totalCitizen total number of citizens totalInt total number of internationals totalY esCitInt total number of citizens and internationals who joined a job in 3 58 totalNoCitInt total number of citizens and internationals who did not join a job in 3 months yesExpCitizen total number of citizens who joined a job within 3 months expected noExpCitizen total number of citizens who did not join a job within 3 months expected noExpInt total number of internationals who did not join a job within 3 months expected yesExpInt total number of internationals who joined a job within 3 months expected
54. e according to your primary key In the Primary Key Field Name combo box select the primary key of your bean Specify an Abstract Schema Name Click Next to skip the remaining dialog boxes until you encounter the Finish button to click Packaging the Web Client WAR file In packaging a Web client the wizard performs the following tasks Creates the Web application deployment descriptor Adds the component files to a WAR file Adds the WAR file to the application s ear file e Select File New Web Component Edit Tools Help Application tan ci Add to Application gt Application Client 8 H lt 42 lt I lt Enterprise Bean ions Newstats Newstats Jar Open Close Resource Adapter Add Server mn Remove Server ejb jar ic jar JAR Display Name NewstatsJar Contents 9 B servers 7 META INF DB ChiSquare class DB ChiSquareBean class DB ChiSquareHome class 3 Correlation class A localhost Edit DB CorrelationBean class 3 CorrelationHome class Hypothesis class D HypothesisBean class Des Manife Icons Deploy Juul 135 e Skip the introduction and Select Create New WAR File In Application If you already have a War file then select Add to existing WAR files A WAR file is required to contains this web component You can create a new WAR fi
55. e getFValue throws java rmi RemoteException Returns the FValue Returns Returns the FValue getA public double getA throws java rmi RemoteException Returns the y intercept Returns Returns the y intercept getB public double getB throws java rmi RemoteException Returns the slope Returns Returns the slope 76 getCoefDeterm public double getCoefDeterm throws java rmi RemoteException Returns the coefficient of determination Returns Returns the coefficient of determination getCoefCorrel public double getCoefCorrel throws java rmi RemoteException Returns the Correlation coefficient Returns Returns the Correlation coefficient getStdErrorEst public double getStdErrorEst throws java rmi RemoteException Returns the standard error Returns Returns the standard error 77 getMSSalaryArray public double getMSSalaryArray throws java rmi RemoteException Returns the salary string for Masters Returns Returns the salary string for Masters getMSGpaArray public double get MSGpaArray throws java rmi RemoteException Returns the GPA string for Masters Returns Returns the GPA string for Masters getPHDSalaryArray public double getPHDSalaryArray throws java rmi RemoteException Returns the salary string for PhD Returns Returns the salary string for PhD 78 getPHDGpaArray public double getPHDGpaArray throws java rmi RemoteException Returns the GPA string for Mast
56. eans are used to maintain the session of a client For example the shopping cart is implemented using the session beans Entity Beans Entity beans model business data such as change the name of a customer modify his address etc These beans represent the data stored in a database They survive the server crashes because it reconstructs the data from the database when the server comes up The Entity beans are used as data access mechanism as they connect to the database and manipulate or retrieve the data depending upon the user reguest The EJB architecture simplifies enterprise applications by basing them on reusable modular components It provides a complete set of services to those components and handles many details of application behavior automatically Project Overview Statistical analysis Tool Session Beans Figure 2 Overview of the Statistical Analysis Tool SAT The Statistical Analysis tool is a web based application which enables the users to register as members and then perform statistical analysis on the member data This website is being developed to provide for the reguirements of the Department of Statistics Kansas State University The statistics department at K State wants their alumni information to be stored and use this information as a basis for future reference The department feels that analysis on data 10 provided by the alumni will help them make deci
57. er to login and displays a message that the login has failed and prompts the user to login again 142 Analysis Department Of Statistics staistis home contactus Alumni Center Log In Failed Please try again On successfully logging into the system the member is directed to the webpage where he can select the analysis he wishes to perform The research questions and the co rresponding test are displayed on the page The member has to click the Analyze button to see the results of any particular test E ir Alumni Home About Alumni Association Membership Alumni News amp Events Services amp Benefits Department of Statistics staistics home contact us Alumni Center Statistical Analysis aaa Select a question for analysis Is there a correlation between the grade point average ofthe alumni and the salary he receives Chi Square Test Does higher GPA result in higher salaries for the Masters or the PhD alumni Is getting a job within three months of graduation independent of the citizenship of the alumni 143 Results Once the member has selected a test the results are displayed in the form of a graphs or tables with all the statistical data The output for each of the tests is given below e Simple Linear Regression Masters Students PHD Students sala
58. ere exists a correlation between grade point average of students and the salaries they obtain The x variable is the Salary and the y variable is the GPA The formulae used for this test Slope b SSxy SSx where SSxy Sum xy Sumx Sum y n and SSx Sum x2 Sum x 22 n y intercept a ybar b xbar where xbar mean of salaries and ybar mean of GPA s Equation of a line y a b x Coefficient of Correlation r SSxy SgRoot SSx SSy Where SSy Sum y2 Sum y 2 n SSx Sum x2 Sum x 2 n and SSxy Sum xy Sumx Sum y n class Correlation operations PerformCtest u Registration 43 ASSOCIATIONS association memberandall between AllRegistrations 0 role membergroup Registration 0 role memberEntity end association servereg between servlet 1 role serve AllRegistrations 0 role servereg end association servlinear between servlet 1 role servel Regression 0 role servelin End association servechisguare between servlet 1 role servec Chisquare 0 role servechi End association servehypo between servlet 1 role serveh Hypothesis 0 role servehyp End association servecorrel between servlet 1 role serveco Correlation 0 role servecorr 44 association test between Registration 0 role testMember Regression 0 1 role testPerform end association hypo between Registration 0
59. ers Returns Returns the GPA string for Masters estimateY public double estimateY double x throws java rmi RemoteException This method calculates the equation of the best fit line Parameters x The x coordinate i e the GPA Returns Returns a b x the equation of line haveData public int haveData throws java rmi RemoteException javax ejb CreateException This method checks to see 1f we have enough data to do the test as we require at least 3 points to calculate standard error of estimate 79 Returns number of points items public int items throws java rmi RemoteException javax ejb CreateException Returns the number of members Returns Returns the number of members itemsMS public int itemsMS throws java rmi RemoteException javax ejb CreateException Returns the number of MS members Returns Returns the number of MS members itemsPHD public int itemsPHD throws java rmi RemoteException 80 javax ejb CreateException Returns the number of PhD members Returns Returns the number of PhD members ejbCreate public void ejbCreate setSessionContext public void setSessionContext javax ejb SessionContext ctx ejbRemove public void ejbRemove ejbActivate public void ejbActivate ejbPassivate public void ejbPassivate ejbLoad 81 public void ejbLoad ejbStore public void ejbStore Class HypothesisBean public class HypothesisBean Hypot
60. esting 22 phase which validates the reguirements of the system and also provides a feasibility study on the project The next phase will be the testing phase which will include testing the complete application after implementation Unit testing functional testing white box testing for the code and the black box testing of the application as a whole unit will be a part of the testing phase Thus the product will be thoroughly tested for each functional reguirement and documented for its reliability 23 CHAPTER 2 PROJECT PLAN Introduction The purpose of the project plan is to give an account of the three phases of the project specifying the iterations and milestones in the development of the Statistical Analysis Tool SAT A Gantt chart is included in the document for this purpose which gives a detail of the plan with estimated dates and sign offs The project essentially spans over three phases of development which are the reguirements phase or the inception phase the design phase or the elaboration phase and the implementation phase or the production phase Inception phase In the first phase of development i e the Reguirement phase a working prototype of the project would be developed to demonstrate the feasibility of the project This would involve gathering reguirements from the end users learning the technology used to develop the application and developing the documentation which includes the reguirement specifica
61. etTotalYesCitInt BSigetTotalNoCitInt BSigetYesExpCitizen BSigetNoExpCitizen BSigetNoE xpint BSgetYesExpint BSigetChiSquareValue Figure 16 Chi Square 98 The detailed method description for this class is as follows getUsers public java util Collection getUsers throws java rmi RemoteException javax ejb CreateException This method gets a collection of all the members in the database from the Registration bean Returns A collection of all the users startChiSquareTest public void startChiSquareTest throws java rmi RemoteException javax ejb CreateException This method initializes all the attribute variables for Chi Square class getObservedValues public void getObservedValues throws java rmi RemoteException javax ejb CreateException Getting the Observed Values for Chi Square test 99 getExpectedValues public void getExpectedValues throws java rmi RemoteException javax ejb CreateException Getting the expected values and chi square value for the table of citizens and international students getDifferences public double getDifferences int obs int exp throws java rmi RemoteException Get the difference of values between the observed and expected value O 2 Parameters obs observed value exp expected value Returns returns difference in the two values get YesObsCitizen public int getYesObsCitizen throws java rmi RemoteException 1
62. eturns A string of salaries makeGPAString public java lang String makeGPAString double x int n This method collects the GPA s of members in an array Parameters x array for GPA n number of items in the array Returns a string of members s GPA registerUser public void registerUser java io PrintWriter out 53 HttpServletReguest reguest HttpServletResponse response throws java io JOException This method registers new members Parameters out PrintWriter object reguest the reguest object for the Servlet response the response object for the Servlet Throws java io IOException the servlet is interrupted is IOException occurs CheckLogin public void CheckLogin java io PrintWriter out HttpServletRequest request HttpServletResponse response throws java io IOException This method does the validation check on members to see if it already exists in the database Parameters out PrintWriter object request the request parameter for the servlet response the response parameter for the servlet Throws 54 java io IOException the servlet is interrupted is IOException occurs calculateCorrelation public void calculateCorrelation java io PrintWriter out throws java io IOException This method gets the results of correlation calculation and displays it to web page Parameters out PrintWriter object Throws java 1o IOException the servlet is interrupted is IOException occurs roun
63. h and the total development time of the statistical analysis tool Reviews and audits Ouality Assurance for this project will include at least one review of all current work in each stage of development of Reguirement Design and Implementation There will be timely reviews by the major advisor of the current work There will be three formal presentations held at the end of each phase of reguirement design and implementation The committee will provide a review after every presentation Test A software test plan will be developed which will outline the test activities reguired for testing Test results will be documented and they will be sufficient to demonstrate that the software reguirements have been met Problem reporting and corrective action The committee will provide reviews on all current work and corrective measures for changes will be taken The errors and problems encountered during the development of the project will be documented Tools technigues and methodologies Oracle Database is used as the database server 113 Sun ONE Application Server 7 is used as a J2EE container which consists of EJB server and Web Server Rational Rose tool is used for the UML model design USE 2 0 1 is used for the formal OCL specification Training The following courses taken by the developer will provide the required training in the development of the project CIS540 SoftwareEngineering I CIS740 SoftwareEngineering II CIS
64. he results will be displayed in the form of graphs or tables All the documents will be refined and a project evaluation will be made An assessment evaluation will be conducted to test the application This phase would be complete when the committee approves the project and the documentation Testing phase The testing involves making test cases to test if the functionality of the project behaves as 25 expected This involves testing all the individual analysis tests and the application as a whole Documentation phase This phase essentially evolved over all phases of the development A final report at the completion of the project would summarize all the artifacts developed during the phases of the project The user manual component design document and the source code with java doc would be included The project plan also provides an updated estimate of the cost size and effort reguired for the implementation of the SAT and an Implementation Plan which will define the activities and actions that have to be accomplished during implementation Gantt Chart D Teak are 9 fih My dre MY 8 S Uo um 47 SK 5 OB WS WO 76 87 34 amt 97 8 180 US 2 18 1 a neepdon Phase Mon don 2 B Devolp OverellReguremento
65. hesisBean is a session bean which handles the computation for the hypothesis test It tests to check the relation between mean salaries of MS and PHD members HypothesisRemote home RegistrationHome coll Collection ESnMS nPhD int 5 meanPhD double E3VarianceMS double ariancePhD double E degrees int Servet pooledSD double Value double po d CDF double qe pValue double lt WSigetUsers void RegistrationBean BSigetMeanSalaries void BSigetVariance void BSigetDistribution double t void BSigetP Value double void BSigetnMS Q int BSigetnPhD int BSigetDegrees int BSigetPValue double BSigetMeanMS double BSigetMeanPhD double BSigetVarianceMS double BSigetVariancePhd double BSigetpooledSd double BSigetTValue double BSigetCDF double Figure 13 Hypotheses The detailed method description for this class is as follows 82 getUsers public void getUsers throws java rmi RemoteException javax ejb CreateException This method connects to the database and gets a collection of all the members in the database from the Registration bean getMeanSalaries public void getMeanSalaries throws java rmi RemoteException javax ejb CreateException This method gets the mean salaries of the Masters and PhD students separately for analysis getVariance public void getVariance
66. his project and the business logic tier includes the Enterprise Java Beans for calculating the research questions The EIS tier consists of the oracle database 12 The technologies that will be used in the project are J2EE reference server as the application server Oracle as the database server Enterprise JavaBeans Entity beans and session beans XML amp XSL Java Servlets Java Script Java Applets and HTML Quality e Scalability The application will be completely scalable as it can be accessed by using a browser Internet Explorer and Netscape Navigator Since the application server used is J2EE which is Java based the application can be installed on different platforms e Reliability The errors will be handled by using java exceptions in the java programs EJB s and Servlets The HTML will include java scripts to handle error checks at the client side Thus both client side and server side errors and exceptions will be handled completely within the application The testing process will include a rigorous process of different types of inputs to the website to check the reliability and rigidity of the application e Security There is no need for a highly secure system for the application It involves the basic security infrastructure of the normal login process for the registered members The members need to register into the application and then login using the login and password that they provided 13 Risks
67. if there is a difference in the average salaries of the Master s student and the PhD students at a 5 level of significance The test is a t test for independent samples The formulae used for this test are t value 1 x2bar s sqroot 1 nl 1 n2 where nl total number of Masters students and n2 total number of PhD students and s sqroot n1 1 512 n2 1 s22 n1 n2 2 812 and 522 are the variances of the salaries for the masters and PhD students respectively 12 sgrootf Sum xi xbar 2 n 1 x bar and 2bar the means of salaries for the masters and PhD students respectively Degrees of freedom n1 n2 2 class Hypothesis operations PerformHtest u Registration end 42 The chisguare class is a session bean which performs the chi sguare test calculations In this test we are testing to see if getting a job within 3 months of graduation is independent of the citizenship of the student at the Masters level at a 5 level of Significance The test is a chi sguare test of independence The formulae used for this test are X2 Sum O E 2 E for each cell where E row total column total sample size degrees of freesom rowtotal 1 columntotal 1 class Chisquare operations PerformChitest u Registration end The Correlation class is a session bean which performs the Correlation test calculations The test checks to see if th
68. ion javax ejb CreateException Start analysis for PhD members addXY public void addXY double x double y throws java rmi RemoteException javax ejb CreateException This method adds the x and y coordinates Parameters x The GPA of the member y The salary of the member 73 Calculate public void Calculate throws java rmi RemoteException javax ejb CreateException This method calculates the slope y intercept coefficient of determination Correlation coefficient mean square due to error fvalue and the standard error for the Masters and PhD members separately getMSError public double getMSError throws java rmi RemoteException Returns the mean square due to error Returns Returns the mean square due to error getMSRegress public double getMSRegress throws java rmi RemoteException Returns the mean square due to regression Returns Returns the mean square due to regression 74 getSSError public double getSSError throws java rmi RemoteException Returns the sum of squares due to error Returns Returns the sum of squares due to error getSSRegress public double getSSRegress throws java rmi RemoteException Returns the sum of squares due to regression Returns Returns the sum of squares due to regression getSSTotal public double getSSTotal throws java rmi RemoteException Returns the sum of squares total Returns Returns the sum of squares total 75 getFValue public doubl
69. is FPA and COCOMO model to predict the development effort of the statistical analysis package Function point analysis and the COCOMO model are an accepted standard for the measurement of software size in software engineering Function Points Analysis The FPA is a reliable method for measuring the size of computer software It essentially measures functionality that the user reguests and receives It also measures the software development and maintenance cost and size independently of the technology used for implementation The general approach that FPA follows is e Count the number of inputs outputs inguiries master files and interfaces reguired then calculate the Unadjusted Function Points UFP e Calculate the adjusted function point AFP by multiplying these counts by an adjustment factor the UFP and the product complexity adjustment e Calculate the Source Lines of Code SLOC with the help of the AFP and the Language Factor LF Calculation of the unadjusted function points UFP 27 The measures functionality that the user reguires The specific user functionality is a measurement of the functionality delivered by the application as for user reguest The 5 function types identified are e external input which receives information from outside the application boundary e external output which presents information of the information system e external enquiries which is special kind of an external output e An ex
70. ize this focus on the learning curve and the optimization of the features set 14 Reguirement specification Product Perspective The SAT is a web based application and hence will reguire a browser like the Internet Explorer 4 0 or Netscape 4 08 The application will be able to connect to remote server and will have an oracle database server It will use J2EE technology for its development The package is independent of any other application The J2EE application will be three tier application with the client tier on the client machine the web tier and the business tier business logic on the J2EE server machine and the database tier composed of the database The business tier or the business logic will be provided by Enterprise Java Beans and servlets and JSP pages will provide the web tier Product functions The members will be able to register online with a login name and password This information is verified with the information in the database and the member is appended The members will be able to add modify or delete the information provided by them when they register The members will be able to select a research question to be analyzed and analysis will be done and the results will be displayed in the form of a graph or table 15 Depending on the research guestion regression hypothesis test chi sguare test or correlation analysis will be performed User characteristics Basic knowledge of using com
71. l be documented This will include class diagrams and object models along with any seguence or interaction diagrams Formal Reguirement specifications A section of the product s design will be formally specified using OCL Test plan A test plan will be developed with appropriate test which will demonstrate that the statistical analysis tool satisfies the reguirements The plan will include test cases and the data that will be used for each case and the results along with the test cases will be documented Formal Technical Inspection A test plan for the formal technical inspection will be selected and a IEEE standard formal checklist will be used for the inspection The two MSE students selected as inspectors are 1 LakshmiKanth Ghanti 2 Divyagyana Nanjaiah The inspectors will provide a report on the result of their inspection and these reports will be documented Fxecutable Architecture Prototype An executable architecture prototype will be built in one or more iterations which will address all critical reguirements identified in the vision document and expose the top technical risks 34 Implementation Plan The Implementation plan will define the tasks to be completed during implementation The tasks are as following User Manual The Manual will describe all the features of the Statistical analysis Tool and the reguirements for the installation of the software It will also describe in detail how to use the tool with sc
72. lIsNew context servlet RegisterUser u Registration pre a u isDefined pre b servereg memberFntity gt excludes u post c servereg memberEntity gt includes u post d servereg memberEntity gt exists ulu oclIsNew context servlet createCorrelation co Correlation pre a servecorr gt excludes co pre b co isDefined post c servecorr servecorr pre gt including co post d servecorr gt exists co co oclIsNew context servlet createHypothesis h Hypothesis pre a servehyp gt excludes h pre b h isDefined post servehyp servehyp pre gt including h post d servehyp gt exists h h oclIsNew context servlet createChiSquare c Chisquare pre a servechi gt excludes c pre b c isDefined 48 post c servechi servechi pre gt including c post d servechi gt exists c c ocllsNew context Regression PerformRtest u Registration pre a u isDefined pre b u testPerform gt isEmpty post c u testPerform gt notEmpty context Correlation PerformCtest u Registration pre a u isDefined pre b u corPerform gt isEmpty post c u corPerform gt notEmpty context Hypothesis PerformHtest u Registration pre a u isDefined pre b u hypPerform gt isEmpty post c u hypPerform gt notEmpty context Chisquare PerformChitest u Registration pre a u isDefined pre b u chiPerform gt isEmpty post c u chiPerform gt notEmpty 49 CHAPTER 4 COMPONENT DESIGN
73. le within an existing application or use an existing WAR file After you have selected a WAR file add the web componentfiles and any other desired files to its contents Optionally you can also provide a description define the advanced settings and choose icons for the WAR file WAR File Location ja Create New WAR File in Application lt gt Newstats E WAR Display Name WebApp1 Add to Existing WAR File Contents CO WEB INF classes I lib C web xml Edit Description Advanced Settings Icons Help Cancel lt Back Next gt e Click Edit e In the tree under Available Files locate the directory where you saved the servlet class or any html pages you wish to add and click Add If you have any java applets that you wish to add you will notice that they get added to the classes where the servlet is Drag the applet class files from there and put them with the html pages e Click OK e Click Next Choose the web component you are creating which is a servlet in our case In the Servlet Filename combo box select one of the servlets you want Click Next Skip and Click Next 136 In the Aliases dialog box Click Add And Enter the alias name for your servlet Click Finish Specifying the JNDI Names and Generating Default SOL 1 Select the entity bean in the JAR file and click on Entity tab File Edit Tools Help
74. messages are displayed if the input is not acceptable Outputs The member is directed to the main page on successful registration Analysis Reguirement Purpose The research question is selected to perform analysis like regression Inputs Input will be the research question selected by the user and consequently the data that the user wants to use for the analysis Processing Depending on the research question the appropriate statistical analysis is performed with the help of the EJB which provide the middle layer in this three tier application It can be regression analysis correlation hypothesis test or the chi square test If an invalid input is entered there will be appropriate error messages handled by using java exceptions in the java programs EJB s and Servlets The HTML will include java scripts to handle error checks at the client side Thus both client side and server side errors and exceptions will be handled completely within the application The SAT will also undergo rigorous testing with various inputs to check whether the analysis is being conducted correctly and that all invalid inputs are not accepted The feasibility of conducting the test will be checked when the member enters a request for a particular test by checking if there is data available in the database for that particular test There have to be at least one Masters and PhD student to do the hypothesis test at 19 least three members to do the
75. mi RemoteException This method gets the critical chisuare value for the test Returns chi sqaure value ejbCreate public void ejbCreate setSessionContext public void setSessionContext javax ejb SessionContext ctx 105 ejbRemove public void ejbRemove ejbActivate public void ejbActivate ejbPassivate public void ejbPassivate ejbLoad public void ejbLoad ejbStore public void ejbStore 106 CHAPTER 5 SOFTWARE QUALITY ASSURANCE Purpose Software Quality Assurance Plan SQAP consists of those procedures techniques and tools used to ensure that a product meets the requirements specified in the software requirements specification This document describes the SQA plan for the Statistical Analysis Tool in partial fulfillment of the requirements of the MSE project The tool essentially performs four kinds of statistical analyses They are Simple Linear regression correlation hypothesis test and Chi square test The scope of this document is to describe all the software processes and documents used for the quality assurance of the statistical analysis tool This plan e identifies the responsibilities of the project developer e lists the activities processes and work products that will be reviewed and audited e Identifies the work products The portions of software life cycle covered by this plan are Software requirements specification System architecture design System detailed design Encodi
76. n about the users and J2EE will be used as an application framework with the help of Enterprise Java Beans to develop the functionality and the business logic of the application Hardware Intel Pentium III 300 MHz or 1 0 GHz Athlon or faster At least 256 MB RAM At least 200 MB freed hard disk space Database Server Use the available oracle database server Software Internet Explorer or Netscape Navigator 17 J2EE Application Server The deploy tool from Sun will be used to maintain the EJB s Functional reguirements Login Reguirement Purpose Provides member authentication Inputs Inputs are through the keyboard and mouse clicks Processing The input is verified by checking if the member already exists in the database Outputs The correct input will result in the next page 1 the analysis page being loaded If the input is incorrect then an error message will be displayed Registration Form Reguirement Purpose Registration of a non member Inputs Inputs are through the keyboard and mouse clicks Processing The input is validated using client side as well as server side validation The client side validation will include checks for missing information in the reguired fields and other text fields like email and phone numbers will be checked for validity The server side validation will involve checking if a member in the database already uses the 18 username entered The appropriate error
77. nent To create a new application 1 Select File New Application In the deploytool 131 Edit Tools Help p N en S Add to Application gt Application Client 8 H E lt Enterprise Bean ions Newstats Resource Adapter Web Context Security Web Component File Name Add Server tatsvVar W Ihome gradsipadmaja 23Nov03 javaZip Newstats ear Remove Server tatsJar Application Display Name Newstats Contents META INF eib jar ic jar D war ic war Add Libra Remove 1 G Description Icons 2 Click Browse 3 Navigate to the directory where the application reside and in the File Name field enter the name you wish to give for the ear file Note that the ear file should be present where all your java files are stored 5 Click New Application 6 Click OK Packaging the Enterprise Bean JAR file In packaging an enterprise bean the wizard performs the following tasks Creates the bean s deployment descriptor Packages the deployment descriptor and the bean s classes in an EJB JAR file Inserts the EJB JAR file into the application s ear file e Select File New Enterprise Bean 132 Select the Create New File when you crating the for the first time If you have already created one and want to add another EJB to thi
78. ng Testing and Integration Users Manual 107 Software guality assurance will be implemented throughout the entire life cycle of the tool s development until the release of the software product Reference documents IEEE guide for software quality assurance planning 730 1 1995 Lecture notes CIS 748 Software Management Dr Scott Deloach Spring 2002 Pressman Roger S Software Engineering A Practitioner s Approach Fifth Edition Mc GrawHill NY June 2001 Management Organization The statistical tool is an individual project and hence there is only one member involved The project has to be supervised by a graduate committee The committee consists of Major Professor Dr Daniel Andresen Committee Dr William J Hankley Dr Scott Deloach Developer Padmaja Havaldar Tasks It is the sole responsibility of the developer to review the tool s efficiency reliability and accuracy The major professor will conduct inspections and reviews The graduate Committee will also review the documents of each phase before every presentation 108 Responsibilities Developer The responsibilities of the developer include Developing all the necessary documents involved in the software life cycle This includes developing the software reguirement specification the SOA plan and cost estimation of the project Developing the design plan of the project Implementing and testing the statistical analysis tool Presenting the documents
79. ns called complexity weighting factors in the table below Each complexity weighting factor is assigned a value complexity adjustment value that ranges between 0 not important to 5 absolutely essential 6 jOnlinedaaentry T Input transaction over multiple 1 screens 8 Master files updated online 3 9 Information domain values complex 5 Total complexity adjustment value 42 Figure 8 Table of complexity adjustment value Calculate the Source Lines of Code SLOC and the formulas used Total Unadjusted Function Points UFP 90 Product Complexity Adjustment PC 0 65 0 01 42 1 07 Total Adjusted Function Points FP UFP PC 96 3 31 Language Factor LF for Java assumed as 38 Source Lines of Code SLOC FP LF 3659 After taking the changes into consideration the source lines of code have changed from 4636 to 3659 COCOMO Model The COCOMO model is a good measure for estimating the number of person months reguired to develop software My project the Statistical analysis package is an application program The table below presents the COCOMO formulae for different types of programs Development Time Programmer Productivit y Month Application PM 2 4 KDSI PM 2 5 PM Programs Utility Programs PM 3 0 KDSI PM 2 5 PM System Programs PM 3 6 KDSI PM 2 5 PM Figure 9 Table of COCOMO formulae
80. nts to calculate standard error of estimate Returns number of points members public int members throws java rmi RemoteException javax ejb CreateException Returns the number of members Returns Returns the number of members ejbCreate 96 public void ejbCreate setSessionContext public void setSessionContext javax ejb SessionContext ctx ejbRemove public void ejbRemove ejbActivate public void ejbActivate ejbPassivate public void ejbPassivate ejbLoad public void ejbLoad ejbStore public void ejbStore 97 Class ChiSquareBean public class ChiSguareBean ChiSquareBean is a session bean which handles the computation for the chi square test It tests to see if getting a job within three months of graduation is independent of the alumni s citizenship ChiRemote home RegistrationHome coll Collection esObsCitizen int noObsCitizen int int EnoObsint int EStotalCitizen int int Semet otalYesCitlnt int E 5 double E3noExpCitizen int esExpCitizen int E int noExpint int BSigetUsers RegistrationBean BSistartChiSquareTest 7 BSgetObservedValues BSigetExpectedValues getDifferences int obs int exp et YesObsCitizen BSigetNoObsCitizen BSigetDegrees BSigetYesObsint BSigetNoObsiInt WSigetTotalCitizen BSigetTotalInt BSig
81. or one J2EE application Web components and their container run on the J2EE server e An application client container manages the execution of all application client components for one J2EE application Application clients and their container run on the client machine e An applet container is the web browser and Java Plug in combination running on the client machine The J2EE framework today provides the standard platform for distributed applications It creates a standard in which application components can be distributed and reused Enterprise Java Beans EJB An enterprise bean is a server side software component for the server or middle tier that can be deployed in a distributed multi tier environment The Enterprise Java Beans allow the implementation of business logic and also the data access implementation Enterprise Java provides built in support for application services such as transactions security and database connectivity The EJB are mainly as follows Sessions Beans Session beans model business processes and process business logic such as logic to compute prices transfer funds between accounts in a bank or perform order entry The two different kinds of Session Beans are Stateless Session Beans These beans do not maintain the state and the information stored in these beans is lost when the server crashes State full Session Beans These beans are also short lived and do not survive the server crashes These b
82. or wrong user name and or empty or wrong password the member should be denied access and be prompted to enter the correct input e If the member enters correct input then the member should be given access and taken to the next page 6 2 Registration Register new member e If the member enters empty or wrong input in the required fields the member should not be allowed to register and a message prompting him to enter correct input should be displayed e Ifthe member fills in all the fields correctly he she should be registered as a new member and the record should be saved in the database e If the member fills in all the fields correctly he she should be allowed to update data and the record should be saved in the database 116 6 3 Analysis performance e The member should be able to select a research guestion from a menu of the four analysis tests The selection should result in another page with the results of the test in the form of a graph or table 6 4 Performance Reguirements e other performance requirements will be tested to see if they actually comply with their reguirements described in section 3 5 of the SRS 7 Item Pass Fail Criteria The testing of SAT will be considered accomplished when all the tests of the mentioned features have been successfully completed 8 Suspension and Resumption Criteria 8 1 Suspension Criteria If any of the tests selected by the member do not give the expected result then
83. othesis 46 inv atleasttwoMembers Hypothesis allInstances forAll h h hypMember gt size gt 2 check if atleat one masters or one phd inv checkforatleastoneMasters self hypMember gt exists m Registration m degree Masters inv checkforatleastonePhd self hypMember gt exists m Registration m degree Phd CHISOUARE CLASS context c Chisguare inv atleasttwoMembers Chisguare allInstances gt forA1l c c chiMember gt size gt 2 This constraint has to check for those members who found a job within 3 months of graduation and then constrain it so that there is at least one citizen and one international inv recentgradscitizen c chiMember gt exists m Registration m joinin3mths true and m citizen true inv recentgradsnoncitizen c chiMember gt exists m Registration m joinin3mths true and m citizen false Operations constraints context AllRegistrations addMember m Registration pre a memberEntity gt excludes m pre b m isDefined post c memberEntity memberEntity pre gt including m post d memberEntity gt exists m m oclIsNew 47 context AllRegistrations DeleteMember m Registration pre a memberEntity gt includes m post b memberEntity memberEntity pre gt excluding m context servlet createRegression r Regression pre a servelin gt excludes r pre b r isDefined post servelin servelin pre gt including r post d servelin gt exists r r oc
84. puters is adeguate to use this application Knowledge of how to use a mouse or keyboard and Internet browser is necessary The user interface will be friendly enough to guide the user Constraints Access to the web is reguired As for the developer constraints the alumni information was not available for security reasons Many assumptions about the data had to be made The learning curve was steep Assumptions and Dependencies It is assumed that alumni data will be made available for the project in some phase of its completion Until the test data will be used for providing the demo for the presentations It is assumed that the user is familiar with an Internet browser and also familiar with handling the keyboard and mouse Since the application is a web based application there is a need for the Internet browser It will be assumed that the users will possess decent Internet connectivity Special Reguirements External interface Reguirements 16 The member has to register using a form provided on the website The user can input data with the help of the keyboard or click with the mouse wherever necessary The package provides pull down menus from which the user can select and links and icons to navigate among the web pages Infrastructure Reguirements Since the tool is web based statistical analysis tool it will use best practice web generation tools and tested technologies Oracle will be used for storing the informatio
85. r the tests without any complexity Since the test is already implemented the developer just has to query to retrieve the proper data from the database and pass it to the tests The Registration bean is an entity bean and creates the table in the database on deployment The member data is stored in the data through this bean Whenever the member selects a research question for a test the servlet creates an instance of that session bean which in turn gets the data from the Registration bean as a collection The Session beans after receiving the data from the RegistrationBean compute the necessary calculations and output the result through XML to the browser 37 The remote classes of the beans define the modules implemented in the bean class and the home classes create and destroy the EJB S RegistrationRemote name String amp bLoginid String double double String amp Degree Boolean amp Citizen Boolean amp joinIn3Mths Boolean Statistical Analysis tool Servlet S CreateRegression double HCreateCorrelation double SgetAllUsers void String ygetGpa Double SgetSalary Double SgetPassword String SgetDegree Char FindAlIQ Collection getCitizenship Boolean SgetloinlraMths Boolean SFindUser String LiD String Pwd Collection o N S CreateHypothe
86. reen shots of the web pages wherever necessary The completion criteria for this task would be when all the features and their use have been successfully described Architecture Design The architecture design will be revised as per the changes reguired and these changes will be documented A detailed component design will also be documented Source Code The source code will be submitted and will be well documented with comments This source code will comply with the architecture described for the implementation of the Statistical Analysis Tool Assessment Evaluation This assessment evaluation will contain a report of the tests done on the SAT and the results of these tests in the form of a test log 35 Project evaluation The project evaluation document will review the process adopted for the implementation of this project and the effectiveness of the methodologies used The completed software will be reviewed to check if it complies with the initial overview of the project The product will also be reviewed to check the guality of the product The implementation of the SAT will be considered completed when e The member will be able to successfully register as a member e The member will be able to successfully login e The member will be able to successfully select an analysis test e regression test will successfully display the correct results e The Correlation test will successfully display the correct results
87. regression analysis at least one citizen and one international student to do the chi square test If these requirements are not satisfied then appropriate error messages will be sent to the member Outputs The output will be a graph or table of the analysis results displayed on the web browser page Behavior Requirements Use Case 1 77 Login a n 2 TU p D hm 2 Register Member Sona Update information e T Select Analysis question Figure3 Member Functions Use Case 2 20 ea I 7 Maintain website d a c A e Update changes System Administrator Add new research questions Figure 4 System Administrator functions Performance Requirements The application should be portable and possible to users of Netscape Navigators well as Internet Explorer Since the application will be displaying graphs for the analysis the response time for a particular analysis should be not be greater than 3 4 seconds for a respectable internet connection speed The database should be scalable it must have the capacity to hold large number of users in future There are about 50 students currently enrolled in the master s and PhD program in the statistics department and the number of enrollment increases every year There are around 10 students graduating from this department every year Assuming that the department uses this website for the next 20 years the alumni should number around 20
88. rning experience and helped to improve the guality of the product 153 CHAPTER 10 FORMAL TECHNICAL INSPECTION CHECKLIST Formal Technical Inspection of the Software Requirements Specification performed by Lakshmikanth Ganti Divyajnana Nanjaiah Criteria Entry Criteria The Inspectors should review the document and evaluate if Yes No Partial and suggest comments if needed Type of Defects to Inspect Completeness Consistency Correctness Feasibility Modifiability Traceability Understandability Maintainability Verifiability Testability Clarity Functionality and Reliability Exit Criteria The inspected document by the inspectors Software Inspection Checklist Completeness Questions Yes No Partial Do the functional requirements cover all No Not Required abnormal situations Are all requirements written at a Yes consistent and appropriate level of detail 5 Does SRS define those reguirements for Partial The SRS does not which future changes are anticipated completely specify this Do the requirements provide an Yes adeguate basis for design Does the SRS include all of the Ye known customer or system needs Is the expected behavior documented Yes Only normal error for all anticipated error conditions conditions are documented 154 Consistency Ouestions Yes No Partial Is there any internal inconsistency No Reguirements are between the software reguirements consistent Does SRS use
89. ry 40 000 3 15 R Sguare 0 305 Equation SE ofh R Square P value 22 3 3 47539 704 36 3 43 120263 528 53065 472 120263 528 53065 472 40091 261 0 305 0 256211 salary 90 000 R Square 925 Equation SE ofh R Square P value 3 85 3 9 3 80052 898 3 85 95 gpa y 76437 2 41692 0x y 76437 2441692 0x 8364 552 0 925 0 037973 SUMMARY OUTPUT SUMMARY OUTPUT Regression Statistics Regression Statistics Multiple R 0 5519 Multiple R 0 962 R Square 0 3046 R Square 0 9255 Standard Error 40091 2608 Standard Error 8364 5519 Observations 6 Observations 4 Intercept 120263 5283 Intercept 76437 2 Slope 53065 4717 Slope 41692 0 ANOVA ANOVA df SS MS df SS MS Regression 1 2 487417453113E8 2 487417453113E8 Regression 1 8 69111432 7 8 69111432E7 Residual 4 5 679159141887 8 1 419789785472E8 Residual 2 6996572 8 3498286 4 Total 5 8 166576595E8 Total 3 9 3907716E7 144 e Hypothesis test sir Department Of Statistics Alumni Center CM Results t Test Paired
90. s file then click on Add to existing Jar file A JAR file is reguired to contain this enterprise bean You can create a new JAR file within an existing application or use an existing JAR file After you have selected a JAR file add the EJB classes and any other desired files to it s contents Optionally you can also provide a description edit the manifest classpath and choose icons for the JAR file JAR File Location i Create New JAR File in Application gt Newstats v JAR Display Name Ejb1 O Add to Existing JAR File Contents 3 META INF Edit Description Manifest Classpat Icons Click on Edit In the tree under Available Files locate the directory where you have stored the files and Add the class files of the remote home and bean classes of the EJB Click OK Click Next In the General dialog box that appears select Session for a session bean and Entity for an entity bean Select the remote home and bean classes as shown below 133 ES aaa 50 si t l Click Next In the Persistence Management dialog box that appears Select the Container managed Persistence 2 0 radio button If in the above figure you select an entity bean then in the Fields to be Persisted check all fields you want in the database Ender Select Methods 134 Enter the Primary Key class Remember to change the typ
91. s for the worst case scenario where the average and standard deviation values shoot up drastically and the throughput gets saturated if the number of users are increased beyond this point SCALABILITY Current server side setup is for small number of simultaneous hits 40 The data is displayed as graphs and tables The data transfer across the wire is minimal as the data is manipulated in the middle tier and the final calculated value is passed over to the web But since the middle tier does a lot of calculations on the data the amount of time required for the result to be displayed is in higher magnitude 129 J2EE application servers are well established and can currently meet an enterprise s applications needs Though they are more complex than the normal web servers they provide a multi tier architecture where each tier can be developed separately and allows the enterprise to add to each tier without affecting the complete application JOEE application servers will also provide some amount of basic clustering facilities e J2EE delivers high performance scalability and reliability for deploying e commerce applications and services e Database scalability Oracle is inherently scalable and can easily port large amounts of data Documentation for Developers Sun ONE Application Server and J2EE A walkthrough for deploying EJB s and the installation procedure for the Sun ONE Application Server is explained below which may
92. sername id else oclEmpty Set Registration endif modifyMember id String name String salary Integer gpa Real Set Registration if memberEntity gt exists m Registration m username id then memberEntity gt select m Registration m name name and m salary salary and m gpa gpa else oclEmpty Set Registration endif verifyldAndPassword id String psw String Boolean memberEntity gt exists m Registration m username id and m password psw end The Regression class is a session bean which performs the linear regression test calculations The test compares to see if having a higher GPA results in higher salaries for the PhD and Masters students The x variable is the Salary and the y variable is the GPA The formulae used for this test are Slope b SSxy SSx where SSxy Sum xy Sumx Sumy n and SSx Sum x2 Sum x 2 n y intercept a ybar b xbar where xbar mean of salaries and 41 ybar mean of GPA s Standard error of estimate Se Square root of SSy b SSxy n 2 where SSy Sum y2 Sum y 2 n Equation of a line y a b x degrees of freesom n 2 Multiple R SSxy SgRoot SSx SSy Coefficient of Determination r2 class Regression operations PerformRtest u Registration end The Hypothesis class is a session bean which performs the hypothesis test calculations In this test we are testing to see
93. ship form is provided by the website to enable the user to register Once a member the user can freely access the data analysis webpage which provides the research guestions pertaining to the four tests The results are displayed in the form of graphs or tables The detailed steps are explained below Log In The member has to log in to perform analysis by e Typing his user name in the UserName field provided at the right hand corner of the webpage e Typing his password in the Password field e Clicking on Submit to gain access to the system Alumni Center Department Of Statistics staistics home contact us Log In for statistical analysis Are you a graduate of the University and the Statistics Department Then you re a Home part of a strong global network of alumni As a member of the Alumni Association you About Alumni pers Association keep touch with the University and other alumni Password Membershi get involved with University enemoetsan enjoy a range of services and benefits BEES get all the latest news from the University Benefits Join now it s free and it s quick and easy to do Alumni News amp Get Involved Alumni News ee Visit the Kansas State Events Do not follow where the path University Alumni Association 4 may lead Go instead where there W is no path and leave a trail 99 me Ralph Waldo Emerson Membership is free e All
94. sions which will benefit future statistics students The alumni data is present at the alumni center at Kansas State University But this information is of a confidential nature and all the information is not available The statistics department also needs some other information for analysis purposes which the k state alumni center does not possess The statistical analysis tool that is being developed has provision for entering this new information that is needed The database will maintain the information about the alumni members and the user can update this information online by logging into his her account The data of all the members will be used as the basis for all the statistical analysis that will be provided on the website The website will allow four types of analysis to be done on the data the regression analysis the chi sguare test the correlation test and the hypothesis test Oracle was chosen as the database for the main reason that it is easily available to the CIS students The project deals with technologies like J2EE Enterprise Java Beans etc and the development of the website using the aforementioned technologies is the main objective of the project Hence the database part of the project was not chosen to be too complicated The tool consists of one table in the database which consists of the member data The table fields are name gpa loginId password salary degree citizenship and join Purpose To develop
95. sis double SCreateChiSquare double S RegisterUser double Hy pothe sisRemote sumY double sumXsguared double amp sumY squared double Tv 1 1 n1 RegressionRemote x double Cerne 0 1 home RegistrationHome coll Collection SyesObsCitizen int home RegistrationHome Call Collection amp nMS nPhD int amp meanMS meanPhD double VarianceMS double BvariancePhD double degrees int amp bpooledSD double amp yiValue double amp CDF double SopValue double CorrelationBean SitemsMS int SitemsPhD int Scalculate void SgetUsers void getMeanSalaries void SgetVariance void getDistribution double t void SgetPValue double cdf void SgetnMS int SgetnPhD int getDegrees int getPValue double SgetMeanMS double SgetMeanPhD double SgetVarianceMS double SgetVariancePhd double SgetpooledSd double getTValue double SgetCDF double amp ysumXY double int strat esObsint int RegeterionHony int 8 5 CorrelationRemote pcosfC double ot otalCitizen int 4 amp bhome RegistrationHome amp stdError double Stotallnt int 0 1 coll Collection amp pe value double Eo otalYesCitint int Salary double Y
96. standard terminology and definitions throughout Is SRS compatible with the operational Not applicable Don t think that the environment of the hardware and project has to define this software Has the impact of software on the Not applicable system and environment been specified Has the impact of the environment on Not applicable the software been specified Do any statements contradict each other Correctness Questions Yes No Partial Is each requirement verifiable by Yes testing demonstration review or analysis Does the SRS identify external Yes Output Variables interfaces in terms of input and output mathematical variables Is there justification for the design implementation constrains Do any requirements conflict with or No Redundancy duplicate other requirements Does the SRS conform to SRS standards 155 Feasibility Ouestions Yes No Partial Will the design operation and Yes maintenance of software be feasible Modifiability Questions Yes No Partial Are requirements organized to allow for he requirements are modifications extensible Is each unique requirement defined more than once Are there any redundant statements Traceability Questions Yes No Partial Can each software functional Yes requirement be traced to a higher level requirement e g system requirement use case Is SRS traceable forward through Yes successive development phases e
97. ternal inguiry presents information of the information system based on a uniguely identifying search criterion without applying additional processing such as calculations e internal logical files contains permanent data that is relevant to the user The information system references and maintains the data and e external interface files also contains permanent data that is relevant to the user The information system references the data but the data is maintained by another information system For each function identified above the function is further classified as simple average or complex and a weight is given to each The sum of the weights guantifies the size of information processing and is referred to as the Unadjusted Function points The table below shows the function types and the weighting factors for the varying complexities 28 Function type Simple Average Complex Internal Logical File 7 10 15 External Interface File 5 7 10 External Input 3 4 6 External Output 4 5 7 External Inquiry 3 4 6 Figure 6 Table of Function types Using these definitions above the files types in my project can be counted as follows For the first presentation I had considered selection of research question for each test as an input But essentially the user can choose only one research question at a time hence the selection of a research question is considered as one single input Similarly for the outputs I had considered the
98. the Latest Are you still paying an annual Click here to view K State fee to the Association Programs more If the user has not registered to be a member he can do so by clicking on the Membership link which will take him to the online membership form The user will have to 141 e Click on the link Membership or Join here e out all the data fields in the form and click Submit e The system will register the user as a member and go back to the main page from where the member can now log in to access statistical analysis by entering his username and password and clicking on the Submit button Note A login form is provided on all the pages of the website to enable the member to perform analysis 12 Department Of Statistics Alumni Center Membership Full Name Login Password Current Salary Grade Point Average Degree Obtained U S Citizen Online Application Form Instructions Please fill out the following form as coi p etely as possible Joined In Three Months staistics home contact us Log In for statistical analysis Password Submit When asked how much educated men were superior to those uneducated Aristotle answered As much as the living are to the dead 5 Diogenes Laertius If the member enters an invalid user name or password the system does not allow the us
99. three months of graduation Returns Number of citizens with jobs within 3 months of graduation getTotalNoCitInt public int getTotalNoCitInt throws java rmi RemoteException This method gets the total number of citizens who did not get jobs within three months of graduation Returns Number of citizens without jobs within 3 months of graduation getYesExpCitizen public int getYesExpCitizen 103 throws java rmi RemoteException This method gets the expected value of citizens who got jobs within three months of graduation Returns Expected value of citizens with jobs within 3 months of graduation getNoExpCitizen public int getNoExpCitizen throws java rmi RemoteException This method gets the expected value of citizens who did not get jobs within three months of graduation Returns Expected value of citizens without jobs within 3 months of graduation getNoExpInt public int getNoExpInt throws java rmi RemoteException This method gets the expected value of alumni who did not get jobs within three months of graduation Returns Expected value of alumni without jobs within 3 months of graduation 104 getYesExpInt public int getYesExpInt throws java rmi RemoteException This method gets the expected value of alumni who got jobs within three months of graduation Returns Expected value of alumni with jobs within 3 months of graduation getChiSquareValue public double getChiSquareV alue throws java r
100. throws java rmi RemoteException javax ejb CreateException This method gets the variance for the salaries and GPA of the Masters and PhD students separately for analysis 83 getSandT public void getSandT throws java rmi RemoteException javax ejb CreateException This method gets the S and T values for the members for analysis getDistribution public void getDistribution double t throws java rmi RemoteException javax ejb CreateException This method calculates the distribution in terms of CDF Parameters t the t value for the distribution getCDF public double getCDF double x This method computes the cumulative distribution function in terms of the beta CDF Parameters x anumber in the domain of the distribution Returns the cumulative distribution at x 84 See Also betaCDF getPValue public void getPValue double cdf throws java rmi RemoteException javax ejb CreateException This method calculates the P value for the distribution Parameters cdf the cdf value for the distribution betaCDF public static double betaCDF double x double a double b The method computes the beta cumulative distribution function Parameters x anumber between 0 and 1 a the left paramter b the right parameter Returns the beta cumulative probability at x 85 logGamma public static double logGamma double x This method computes the log of the gamma function Parameters X a positive number Ret
101. tion plan software guality assurance plan architecture elaboration plan and the cost estimation plan At the completion of phase I a presentation summarizes all the above processes Elaboration phase The second phase of development consists of the design of the project An object model specifying the architecture design would be developed The entity bean and two session beans will be developed which execute two of the statistical tests namely the simple 24 linear regression and the correlation bean A design of the website will be also be developed The reguirements will be formally specified using the USE tool and object constraint language would be used to write the constraints Two inspectors would conduct a formal technical inspection of the reguirement specification and the results of the inspection will be documented and produced along with all final documents A test plan would be developed and a second executable prototype will be implemented for demonstration at the second presentation All the changes suggested during the first presentation will be considered and the documents will be refined The design phase would be complete when the committee members approve the documents and the progress in the project Production phase This phase consists of the complete implementation of the project The four statistical tests which are four session beans and the registration bean which is an entity bean will be implemented T
102. urns the log of the gamma function at x getnMS public int getnMS throws java rmi RemoteException Returns the number of MS members Returns returns number of MS members getnPHD public int getnPHD throws java rmi RemoteException Returns the number of PHD members Returns returns number of PHD members 86 getDegrees public int getDegrees throws java rmi RemoteException Returns the degrees of freedom Returns returns degrees of freedom getPValue public double getPValue throws java rmi RemoteException Returns the P value Returns returns the P value getMeanMS public double getMeanMS throws java rmi RemoteException Returns the mean for MS members salaries Returns returns the mean for MS members salaries 87 getMeanPHD public double getMeanPHD throws java rmi RemoteException Returns the mean for PHDmembers salaries Returns returns the mean for PHD members salaries getVarianceMS public double getVarianceMS throws java rmi RemoteException Returns the variance for MS members salaries Returns returns the variance for MS members salaries getVariancePHD public double getVariancePHD throws java rmi RemoteException Returns the variance for PHD members salaries Returns returns the variance for PHD members salaries 88 getPooledSD public double getPooledSD throws java rmi RemoteException Returns the pooled standard deviation Returns returns the pooled standar
103. vided on the website which enables the alumni to register The test cases used for this form are given in table 18 The table also includes test cases to check the functionality of the analysis tests in the statistical analysis tool 119 Test Unit Login Password Name Salary Gpa Empty login name Empty password Empty Name Empty salary Non numeric salary value Non positive salary value Empty Gpa Non numeric Gpa value Gpa gt 4 or lt 0 Correct data in each field Prompts the member to enter valid login name Message Please enter a valid login name Prompts the member to enter valid password Message Please enter a valid password Prompts the member to enter valid name Message Please enter a valid name Prompts the member to enter a salary value Prompts the member to enter a numeric salary value Prompts the member to enter a positive salary value Message Please enter a positive numeric value Prompts the member to enter value Prompts the member to enter a numeric Gpa value Prompts the member to enter a value between 0 to 4 Message Please enter a number between 0 and 4 The member will be taken to the main page to log in after he enters all correct values Once the member successfully logs for statistical analysis he is directed to the questions page where he can begin the analysis Regression Bean Less than 3 members No MS members No PhD members Message Not
104. write J2EE applications as the business logic is organized into reusable components and the developers can concentrate on writing just the business logic rather than the low level details while the J2EE platform handles all the underlying services J2EE Containers provide the reguired interface between a component and the low level functionalities specific to a platform that support the component We need to assemble the web tier components and business logic components into a J2EE application and deploy then into containers Container settings customize the underlying support provided by the J2EE server which includes services such as security transaction management Java Naming and Directory Interface JNDI lookups and remote connectivity J2EE components are packaged separately and bundled into a J2EE application for deployment A J2EE application is composed of one or more enterprise bean web or application client component modules A J2EE application with all of its modules is delivered in an Enterprise ARchive EAR file The deployment process installs J2EE application components in the following types of J2EE containers The J2EE containers are described as follows e Enterprise JavaBeans EJB container manages the execution of all enterprise beans for one J2EE application Enterprise beans and their container run on the J2EE server e web container manages the execution of all JSP page and servlet components f
105. y The resultant output was compared with the excel output for known values from table 1 Mention must be made that the tests were tested using different sets of data The above table is one of the data sets In one of these data sets there were egual number of Masters and PhD students and the results came out right But when different number of MS and PhD students were used the result for PhD did not compare with the excel output This error was debugged and the results were comparable The error was due to the fact that the loop for the PhD students used MS size instead of PhD Known values Excel output For MS students SUMMARY OUTPUT Regression Statistics Multiple R 0 55189228 R Sguare 0 304585088 Observations 6 ANOVA df SS MS Regression 1 248741745 3 248741745 3 Residual 4 567915914 2 141978978 5 Total 5 816657659 5 123 Coefficients Standard Error Intercept 120263 5283 137064 8226 X Variable 1 53065 4717 40091 26078 Observed Values using tool SUMMARY OUTPUT Regression Statistics Multiple R 0 5519 R Sguare 0 3046 Standard Error 40091 2608 Observations 6 Intercept 120263 5283 Slope 53065 4717 ANOVA df SS MS Regression 2 487417453113 8 2 487417453113 8 Residual 4 5 679159141887 8 1 419789785472 8 Total 5 8 166576595E8 Figure 19 Table of Regression Result Result of test The tool was successful in computing the test For PhD students Known values Excel output SUMMARYOUTPUT 1

- CiteSeer

Contents

Download Pdf Manuals

Related Search

Related Contents