Hi,
I was going through some old numstar code, trying to get it to work with
scipy core, and I was initially frustrated by the lack of "UInt8" in
scipy, because it's the only name for this dtype shared between numarray
and Numeric. To investigate the situation further, I wrote a little
script (attached), which I think should be used to add a few more names
to the scipy namespace in the name of backwards compatibility with
numstar. Here's the output on my machine:
scipy Numeric numarray
UInt8 - + +
UnsignedInt8 + + -
uint8 + - +
UInt16 - + +
UnsignedInt16 + + -
uint16 + - +
UInt32 - + +
UnsignedInt32 - + -
uint32 + - +
UInt64 - - +
UnsignedInt64 - - -
uint64 + - +
uint + - -
uint_ - - -
Int8 + + +
int8 + - +
Int16 + + +
int16 + - +
Int32 + + +
int32 + - +
Int64 - + +
int64 + - +
int_ + - -
Float32 + + +
float32 + - +
Float64 + + +
float64 + - +
Float128 + - -
float128 + - -
Float + + +
float - - -
float_ + - -
complex64 + - +
Complex64 + + +
complex128 + - +
Complex128 + - -
complex256 + - -
Complex256 - - -
Complex_ - - -
complex - - -
complex_ + - -
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dtypes.py
Type: text/x-python
Size: 1252 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20051120/a4895bfe/attachment.py>