@brandtbucher : While arraykit is passing all SF tests on older NumPy, with NumPy 1.20.1 we are getting the following error. Is this something we can fix in the arraykit implementation?
self._row_dtype = resolve_dtype_iter(b.dtype for b in self._blocks)
TypeError: argument must be an iterable over numpy.dtype, not numpy.dtype[str_]
@brandtbucher : While
arraykit
is passing all SF tests on older NumPy, with NumPy 1.20.1 we are getting the following error. Is this something we can fix in thearraykit
implementation?https://github.com/InvestmentSystems/static-frame/runs/2386175564?check_suite_focus=true