Randomization test with constraints

  • My current project is written in C++. It takes a certain config file to do various work. There are a large number of parameters inside the config. I would like to do randomization test for those inputs. What I would like to achieve is following:

    1. For each parameters, I do uniformly randomization with its own constraints.
    2. The whole testing sample would be drawn uniformly from whole parameter space.

    I am new to the SW testing. But what I would like to achieve is something like UVM does for HW verification. I am wondering if there is any framework for this. Python is preferred.


  • The first thing that came to my mind was Spotify's Model-Based Testing, it doesn't describe your case exactly but looks like you can use a similar method.

    Build a model of your constraints and let a tool run through the combinations.

    Also try and evaluate Microsoft's PICT tool as the source for your constraints model description, it has the added benefit that if some parameters are known to be independent of each other the tool will help you decrease the number of needed tests

Suggested Topics