[Numpy-discussion] KeyError with float96 in linalg.py: polyfit

[Jan-Matthias B. <jan...@gm...>]

Hi all,

I'm testing some computations with float96 at the moment and right now I ha=
ve=20
problems with polyfit raising a KeyError for the keycode 'g', which is=20
floatxx with xx>64.=20

I am getting a KeyError using polyfit on some float96 values. The used=20
Routines seem to know nothing about this type.

My main question is: have I missed something? Shouldn't this type be used?
Below is a more detailed descripton.

Thanks in advance,
	Jan


=2D------------------------------------------------------------------------=
=2D--


In file numpy/linalg/linalg.py, the following definitions at lines 26ff see=
m=20
to be the offending ones:
	# Helper routines
	_array_kind =3D {'i':0, 'l': 0, 'q': 0, 'f': 0, 'd': 0, 'F': 1, 'D': 1}
	_array_precision =3D {'i': 1, 'l': 1, 'q': 1, 'f': 0, 'd': 1, 'F': 0, 'D':=
 1}
	_array_type =3D [['f', 'd'], ['F', 'D']]

Here the new typecodes are missing. I tried=20
	# Helper routines
	_array_kind =3D {'i':0, 'l': 0, 'q': 0, 'f': 0, 'd': 0, 'g': '0', 'F': 1,=
=20
			'D':1, 'G':1}
	_array_precision =3D {'i': 1, 'l': 1, 'q': 1, 'f': 0, 'd': 1, 'g': 1,=20
			'F': 0, 'D': 1, 'G': 1}
	_array_type =3D [['f', 'd', 'g'], ['F', 'D', 'G']]

which gets me a step further to a TypeError:

  File "lib/python2.3/site-packages/numpy/linalg/linalg.py", line 454, in=20
lstsq
    bstar[:b.shape[0],:n_rhs] =3D b.copy()
TypeError: array cannot be safely cast to required type


(Question: Why only one typecode for a type which varies in bitlength on=20
different platforms? On Opteron CPU's I've seen float128 with 'g'?)