C
Colin J. Williams
These timing results with Fibbonacci, using and not using memoize might
be of interest.
Version 2.3.3 - Unwrapped is using interation in place of recursion
C:\Test>python tFib.py
Unwrapped, using fib: 7.52 usec/pass
Raw Fibonacci: 252.66 usec/pass
Fibonacci with old deco: 3.18 usec/pass
Version 2.4.2a
C:\Test>\Python24\python.exe tFib.py
Unwrapped, using fib: 4.92 usec/pass
Fibonacci, using @memoize: 2.89 usec/pass
Raw Fibonacci: 214.11 usec/pass
Fibonacci with old deco: 2.90 usec/pass
Conclusions:
2.4.2a provides a significant speedup for these cases.
The unwrapped approach is much better than recursive fibonacci, but not
as good as with memoise.
This is based on Dan Bishop's code.
Colin W.
be of interest.
Version 2.3.3 - Unwrapped is using interation in place of recursion
C:\Test>python tFib.py
Unwrapped, using fib: 7.52 usec/pass
Raw Fibonacci: 252.66 usec/pass
Fibonacci with old deco: 3.18 usec/pass
Version 2.4.2a
C:\Test>\Python24\python.exe tFib.py
Unwrapped, using fib: 4.92 usec/pass
Fibonacci, using @memoize: 2.89 usec/pass
Raw Fibonacci: 214.11 usec/pass
Fibonacci with old deco: 2.90 usec/pass
Conclusions:
2.4.2a provides a significant speedup for these cases.
The unwrapped approach is much better than recursive fibonacci, but not
as good as with memoise.
This is based on Dan Bishop's code.
Colin W.