Please enable JavaScript.

Coggle requires JavaScript to display documents.

Estimating Software size - Coggle Diagram

- - - - Inputs,outputs, Inquires, data files and interface
- - - - Proxy is a substitute or a stand in, it can help to to judge the product size, they are objects, screens, files, scripts or function points
    - - Related to development effort.
        
        The effectiveness of a proxy is determined by obtaining historical data on a number of products develop. By estimating the size of the proxy you can determine the size of the job.
      - Automatically countable proxy content.
        
        The data must be automatically countable, so it means that the proxy should be a physical entity that can be defined and identify.
      - Easily visualized at projects beginning.
        
        The usefulness of a proxy depends on the degree to wich it helps to visualize the size of the project to be done. There would not be a best proxy for all purposes.
      - Customizable to the needs of the using organization.
        
        Is important to gather and use data that are relevant to a particular project being estimated.
      - Sensitive to implementation variations.
        
        The proxies that are easily identify are app entities such as inputs, outputs, files, screens and reports. Is essential that the languages, design, styles, and application categories to be used are represented in the data used to calculate estimating factors.
    - - Objects could be a good estimating proxies, to determine if they are good size proxies you should examine historical data, so your highest level product could be visualized during requirement analysis.
      - In order to use them, first divide historical object data into categories and size ranges. Object size size is a matter of style,
- - - - It gives you an aproach to the expected product objects and their functions, here you postulate the objects that would be needed
    - - Here you find the objects in the database that most closely resembles, for a new object you judge how its size compare to the when the BD in its category.
    - - Here you determine the size of the based program you are enhancing and their changes. If you find objects or procedures required for your conceltual design you could reuse them
    - - The linear regression method produces two parameters that you can use to calculate the estimated program size.
    - - The purpose is to obtain the bet fit of a straight line to the historical data that have minimum variance.
    - - Using historical data and t distribution, helps you calculate the prediction interval, it is the estimated value from the program LOC minus the interval.