Re: [Cython] FW: cython and hash tables / dictionary

Robert Bradshaw wrote:
> On Wed, 2 Sep 2009, Christopher Barker wrote:
> 
>> Sanne Korzec wrote:
>>> The main bottleneck in my code is a large dictionary / hash table which
>>> I would like to optimize.
>> In what way do you need to optimize it? i.e. how is it used? do you have
>> memory issues or speed issues? python dicts are highly optimized
>> already, so you're not likely to do much better with the look-up speed.
> 
> It could help with both memory and speed. In particular, to do a lookup in 
> a Python hashtable you need to
> 
> (1) Wrap the float in a Python object
> (2) call __hash__ on that new object

I hope that isn't what actually happens  Floating point and hashes 
don't seem like a good idea.

(Just a note, I believe the OP was talking about floats in the values so 
we're good.)

> (3) actually do the lookup
> (4) unwrap the result back into a float.
> 
> Python does have a highly optimized (3), but the overhead for the rest 
> will probably overwhelm it speedwise, so I bet a simple, unwrapped 
> implementation would still be quite a win.
> 
> - Robert
> _______________________________________________
> Cython-dev mailing list
> Cython-dev@...
> http://codespeak.net/mailman/listinfo/cython-dev

--

-- 
Dag Sverre