Regression tests run insane with a set of arguments and compare the
output with a reference. This allows to assess that the output did not
change between the version of insane with which the reference has been
generated and the current version of insane. It does not test that the
result is valid!
Run tests/test_regression.py as a script to generate the reference
files. Run it with pytest or nosetest to execute the tests.
This pull request depends on #2 being merged. Indeed, being able to set the random seed is crucial to be able to reproduce the outputs.
Regression tests run insane with a set of arguments and compare the output with a reference. This allows to assess that the output did not change between the version of insane with which the reference has been generated and the current version of insane. It does not test that the result is valid!
Run
tests/test_regression.pyas a script to generate the reference files. Run it with pytest or nosetest to execute the tests.This pull request depends on #2 being merged. Indeed, being able to set the random seed is crucial to be able to reproduce the outputs.
EDIT:
This that are still to do:
-f,-p, and-m)