I think I had written out the linear algebra previously, and we migrated to the more readable np.linalg.lstsq. However the output of that function (for numpy '1.21.6' at least) is a tuple of properties. The zeroth element of that tuple is what we desire, so accessing coeffs[0] should fix the problem.
The
.fit_continuum()
method appears to be broken on line 393-395:I think I had written out the linear algebra previously, and we migrated to the more readable
np.linalg.lstsq
. However the output of that function (for numpy'1.21.6'
at least) is a tuple of properties. The zeroth element of that tuple is what we desire, so accessingcoeffs[0]
should fix the problem.