search

shunwang / numexpr

Automatically exported from code.google.com/p/numexpr
MIT License
0 stars 0 forks source link

int8 or int16 support #106

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
Most of our numpy arrays are int8 or int16.
When will numexpr support int8 and int16 types?

Currently, the following code does not work:
aa=np.zeros(shape=10000000, dtype=np.byte)
ne.evaluate("where(aa=1,2,0)", out=aa)

Error message:
TypeError: Iterator requested dtype could not be cast from dtype('int32') to 
dtype('int8'), the operand 0 dtype, according to the rule 'safe'

Original issue reported on code.google.com by johnqdel...@gmail.com on 10 Jun 2013 at 6:49

GoogleCodeExporter commented 9 years ago

Original comment by gdemen...@gmail.com on 12 Jun 2013 at 12:30