Improve your Python development experience with flake8-numba. This plugin integrates with Flake8 to provide comprehensive code analysis for projects that make use of numba. When working with numba, many errors are not discovered until the code is run. However, many of these issues can be perfectly caught by performing some basic syntactical analysis. This tool helps you catch potential errors and enhance code quality effortlessly.
Make sure that you have flake8
installed. Then:
pip install flake8-numba
After it calling flake8
will include all rules defined by this plugin.
Some examples are:
@vectorize([float64(float64, float64)])
def f(x, y):
return x + y, 2 # ERROR: only 1 value can be returned
or:
# ERROR: Dimensions mismatch (second argument at left is an array but an scalar at right)
@guvectorize([(float32, float32[:], float32)], "(), () -> ()")
def func(...) -> None:
...
All available rules can be read in RULES.md