- Python 3.7.2
- ply 3.11
- numpy 1.16.2
- scipy 1.2.1
python opt.py [N] | python interpreter.py [N]
- run FFT
python opt.py [N] | python cost.py
- show calcuration cost
python test_opt.py
- run unit test
I implemented Cooley-Tukey FFT and Rader's algorithm.
- Sum of cost from N = 2 to 256 (calculated by
cost.py
)naive.py
(original): 16875645opt.py
: 3564224 (21.1%)
- max error
- 8.83390668259562e−13 (when N=212, initializer=sin)
Unknown, beucase this repository includes source code base on https://github.com/pfnet/intern-coding-tasks
I don't encourage you to include these source codes into other program.