I am working on my documentation of final year project (cardioVascular risk assessment using machine learning model). COuld any one please tell me the difference among dfd level 0, level 1 and level 2. Thanks in advance
The principle of top-down analysis is deeply rooted in structured analysis and design methods in general, and data-flow diagrams in particular.
The principle is that a top-level diagram shows the data-flows between processes in the scope and the external entities. Every process can be exploded/decomposed into an own diagram with more detailed processes and flows, and so on, until the primitive level is reached, i.e. a process which is so elementary that it is not desirable to explode it further.
There are some slight differences in the leveling terminology between the main authors:
- Gane & Sarson use the term “exploding into lower-level”. They mention a top-level diagram showing the data-flows with the main processes of the system. They recommend the use of decimal notation in the decomposition. So at top level, you’d have 1, 2, 3. In the diagram of the next level for process 2, il will be 2.1, 2.2, 2.3, etc. They ate not very explicit about level numbering.
- DeMarco/Yourdon uses the term “partitioning”. He defines precise conventions for the leveling: the top level diagram is called level 0. The successive levels are incremented. The idea is that the level number indicates the number of components in the numbering prefix (level 0 has no prefix: you’ll have process 1,2,3; level 1 has 1 number in the prefix, for example 2 for process 2.1, 2.2, 2.3, level 2 has 2 components as prefix, for example for the diagram of 2.3, you’ll have 2.3.1, etc). Demarco also adds a context diagram on the top of the top level that shows the dataflows of the system in scope with the external entities. To add confusion, sometimes the context diagram is called level zero which is not to be confused with level 0 (sic!).
In your case the level 0 would in principle be the main processes in your system. The set of level 1 diagrams would be, for each top-level process, an own diagram that further breaks down that process into subprocesses with their dataflow.