DFD TUTORIAL

Data Flow Diagram Tutorial

Objectives

After completion of study of this unit you should be able to:

• Describe the use of data flow diagrams

• Produce a data flow diagram from a given case study including different levels

• Distinguish between the different categories of data flow diagrams

1. Introduction

1.1 CASE tools

1.2 Development and purpose of Data Flow Diagrams

2. Components of Data Flow Diagrams

2.1 Components

2.2 Hints on drawing

2.3 Data flows

3. Developing Data Flow Diagrams

3.1 Introduction

3.2 Context diagram

3.3 Level 1 Data Flow Diagram

3.4 Lower levels of Data Flow Diagrams

3.5 Check list

4. Categories of Data Flow Diagrams

4.1 Physical

4.2 Logical

4.3 The relationship between logical and physical Data Flow Diagrams

5. An example of the development of a Data Flow Diagram

6. Process models

7. Summary

8. Activities

9. Bibliography

10. Commentary on activities

1. Introduction

This unit deals with one of the major techniques for recording the requirements of a user for a new computer application. An initial diagram is constructed to show the processes which are being implemented in an existing system. The diagram helps to show how information is used to produce the functions that are required by the current system. It also shows what information is provided to the system and what information is provided form the system. Other benefits include the documentation of who is using the system and what data will be stored. By careful construction of the DFDs (data flow diagrams) the boundaries of the system to be built may be clearly identified. This helps to clarify what will and what will not be constructed. It will also show the interaction that may be required with other systems.

The data flow diagrams should also have some associated documentation. This is necessary as the diagrams are meant as a visual representation of the way in which information is processed. There is limited space on the diagrams so that documentation to explain, refine and describe further details of what is shown need to be kept somewhere in the proposed system documentation. The data flow diagrams and the associated documentation together combine to form a data flow model. This is also commonly called a process model.

The user requirements when complete are used as a basis for the development of the system. Later when the system has been developed it can be tested against the initial requirements to see whether the user’s needs have been met.

1.1 CASE tools

For many ways of developing and implementing new systems, software is available to help the systems analyst produce what is required. This software is called a CASE tool. CASE stands for Computer Aided Software Engineering. Data Flow Diagrams are usually produced using a CASE tool although they can be produced simply with a pencil and paper. The diagrams shown in this unit were developed using SELECT SSADM Professional version 4.1.1.

Using a CASE tool for construction of the DFDs has many advantages. It is not just a drawing tool. Firstly the CASE tool will not allow a non-standard use of notation for all the items in the diagrams. Secondly it applies some rules so that designing the diagrams prevents the users from making connections between the different items that should not be allowed. There is also some checking that may be invoked to alert the designer to some potential errors. As we will see the CASE tool helps to check that the DFDs are consistent with other views of the system which require diagrams that may also be drawn with the CASE tool.

You would find it helpful to become familiar with a CASE tool to practice some of the example activities in this unit.

1.2 Development and purpose of DFDs

Before attempting to construct an initial DFD it is necessary to gather and digest information that helps us to understand how data is processed in the current system. Fact-finding techniques are used for this purpose and are discussed in another Unit. As the DFD is constructed a systems analyst will often come across areas of doubt where the precise way to model the system is unclear. This is a natural part of the development and should not be regarded with alarm. In fact, it is expected and it is a consequence of attempting to model the current situation that questions will be asked to clarify the exact processes which are taking place. Sometimes the analyst will make an assumption and then check this with the user at a subsequent meeting.

Results of interviews, documents, reports, questionnaires etc. will all play a part in helping the analyst to gain an insight into the current processes. Where a system is being developed form scratch the analyst will work with the user to develop the proposed DFDs.

When all the information about the current system is gather it should be possible to construct the DFDs to show:

• the information that enters and leaves the system

• the people/roles/other systems who generate and/or receive that information

• the processes that occur in the system to manipulate the information

• the information that is stored in the system

• the boundary of the system indicating what is (and what is not) included

As a staring point, it is sometimes useful to construct a document flow diagram. This diagram shows how ‘documents’ are passed around an organisation to fulfil the current requirements of

...