Data-flow Diagrams

1
Data-flow Diagrams

• Chapter 17

2
Elements of data-flow diagrams

• There are four basic elements of data flow
  diagrams
• Data Transforms where data is filtered,
  formatted or synthesized.
• Information sources and sinks essentially
  inputs and outputs to the system.
• Information stores where information is kept
  for the duration of system activity.
• Information Flows data that is passed into and
  out of transforms, information stores.

3
Information flows

• Information flows represent the information being
  passed into or out of a transform, source, sink
  or store. It is represented as a labeled arrow

Taxable Income
4
Data transforms

• Data transforms labeled circles with one or more
  incoming and outgoing information flows

Taxable Income
Tax owed
Calculate Tax
Tax Tables
5
Information sources and sinks

• Information sources and sinks are information
  that come into the system, or leave the system,
  and are represented by squares

Pay Stub
Hours
Payroll System
Pay Check
Rate
6
Information stores

• Information Stores represent locations where
  information can be store for the duration of the
  system activity

Total Payroll Expense
7
Refining DFDs

• DFDs are refined (decomposed) to the point that
  they contain only primitive transforms. They
  are refined by drawing another DFD as if they
  were a system in themselves!
• A primitive transform has only one input flow
  or only one output flow (note that it can have
  multiple input flows or multiple output flows,
  but not both!)

8
Heritage numbers

• Refinement of DFDs leads to a problem in larger
  systems we have a large number of diagrams!
  When we look at a diagram by itself the only way
  we can tell what it represents is by looking at
  the parent diagram.
• Heritage numbers allow this. A data transform
  numbered 2.4.3 is the child of DFD number 2.4
  which is the child of DFD 2 in the level 1 DFD
  diagram.

9
The data dictionary

• As the data is defined in the DFD (the flows),
  definitions should be entered in the data
  dictionary.
• The data dictionary is a reference that allows
  us to determine the type of the named data (int,
  float, struct, whatever.)
• This will be essential during the programming
  phase, when knowing what variable names to use.

10
Checking DFDs - CASE tools

• Since DFDs are a formal method, tools have been
  developed that allow automated drawing of the
  DFDs, and management of refined DFDs.
• Since this can be automated, checking DFDs can
  also be automated (such as Process Analyst).

11
Structure charts

• Translation of a DFD to a structure chart is a
  mechanical procedure (see the book.)
• The procedure involves isolating the central
  transform area, which is independent of the
  format transforms and the validation transforms.
• Structure charts also provide a schematic showing
  the flow of the data in a control diagram!
• Now we have a complete picture DFDs show how
  the data flows, and structure charts show how
  control is passed!

12
Payroll System
Format Paycheck
Calculate Withholding
Format Accounting
Calculate Gross Pay
Validate Timecard
Input/output stream At highest abstraction
Central transform
13
(No Transcript)
14
Implementing a DFD

• Once we have a structure chart, the actual
  implementation also becomes mechanical. An
  essential pre-defined form becomes apparent for
  implementing structure charts. An example using
  C is available in the book.