increase test coverage
a CI test runner would also be nice.
We are not aiming for 100% coverage of everything here, but at lest the components core
and flagger
should be covered extensively. All test functions should also be tested, but in a way that at least assures, that they actually run with the passed input. A full check of their performance within the quality control process is out scope here.
python -m pytest --cov=saqc .
reports the following:
-
core
:-
core.py
: 83% -
evaluator.py
: 97% -
reader.py
: 90%
-
-
flagger
:-
baseflagger.py
: 94% -
categoricalflagger.py
: 96% -
dmpflagger.py
: 99% -
simpleflagger.py
: 100% -
continuousflagger.py
: 96%
-
-
funcs
:-
break_detection.py
: 91% -
spike_detection
: 91% -
functions.py
: 93% -
harm_functions.py
: 87% -
machine_learning.py
: 100% -
constants_detection.py
: 96% -
soil_moisture_tests.py
: 90% -
register.py
: 100%
-
-
lib
:-
types.py
: 100% -
tools.py
: 68% (quite a few of the functions are currently not used) -
statistic_functions.py
: 40% (quite a few of the functions are currently not used) -
plotting
: 30%
-