Data Flow Diagram-1 (basic composition of DFD)

Data Flow Diagram (DFD) is one of the most common structural analysis tools, graphically depicts the data movement in the system.

The data flow diagram has four basic components, as shown in Figure 1.3.

Figure 1.3 Basic components and symbols of a data flow chart

Data Flow indicates the flow of data. Processing indicates the processing process of data. Its name should be concise and concise to indicate what processing is completed; data storage is used to store data in a data flow diagram. Data streams that point to data storage can be understood as writing data, while data streams derived from data storage can be considered as reading data, two-way data flow can be understood as modifying data. The data source point or end point indicates that the data source point or end point appears in the graph, and its appearance in the graph is just a symbol, it does not need to be designed and implemented in the form of software.

In a data flow diagram, if more than two Data Streams direct to one processing or are derived from one processing, there is a certain relationship between these data streams. We usually use the symbol shown in Figure 1.4 to represent this relationship.

Figure 1.4 representation of the relationship between multiple data streams and Processing

When drawing a data flow diagram, the following issues are worth noting:

1) Draw a Data Flow Diagram Instead of a program diagram.

Many people are familiar with computer language learning.ProgramIt is easy to mix data flow diagrams. The program diagram describes the system from the perspective of Data Processing. Its arrow is the control flow, indicating the order of data processing. It is used to describe how to solve the problem; a Data Flow chart describes the system from the data perspective. Its arrows are data flows, indicating the flow direction of data. It is used to describe the problem.

2) data stream and processing naming

Generally, the data stream is first named and then processed. When naming a data stream, you should avoid names with no specific meanings such as "data" and "input". When naming a data stream, the ideal practice is to take the name composed of a specific thing verb and a specific object. If two verbs are required, you can consider splitting the processing into two. When naming, the name should be suitable for the entire data stream or processing, rather than simply reflecting some of its components. If you find that a data stream or processing is difficult to name, it is likely that the data stream graph is improperly decomposed. In this case, you should consider re-decomposing the data stream graph.