/numfi

a subclass of numpy.ndarray that does fixed-point arithmetic

Primary LanguageJupyter NotebookMIT LicenseMIT

numfi is a numpy.ndarray subclass that does fixed-point arithmetic.

Feature:

  • Automatically perform fixed-point arithmetic through overloaded operators

  • Maximum compatibility with numpy and other library, just like a normal numpy.ndarray

  • Optimized calculation speed by minimizing quantization as much as possible

Install

Prerequisite: python3 and numpy

pip install numfi

or you can just copy numfi.py and do whatever you want, after all it's only 200 lines of code

Quick start

from numfi import numfi

# numfi(array=[], signed=1, bits_word=32, bits_frac=16, rounding='round', overflow='wrap')
x = numfi(np.random.rand(3),1,16,8) 
# numfi.__repr__() return brief description of numfi object: x => s16/8-r/s
# s for 'signed', followed by word bits and fraction bits, r/s for 'round' and 'saturate` for rounding/overflow method

# any arithmetic operation with numfi will return a numfi object with proper precision and value
# By overloading operators, numfi object can do fixed-point arithmetic easily:
y = x + 1
y = [1] - x
y = x * np.random.rand(3)
y = numfi([1,0,0.1234],1,21,15) / x
y = -x
y = x ** 0.5
y = x % 3
y = x & 0b101
y = x | 0b100
y = x ^ 0b001
y = x << 4
y = x >> 2
y = x > 0.5
y = x >= 0.5
y = x == x
y = x <= np.ones(3)
y = x < [1,1,1]
...

# By inheriting from numpy.ndarray, numfi object can be used just like normal numpy array, and return same numfi object back
y = np.sin(x)
y = x[x>1]
y = x.sum()
y = x.reshape(3,1)
plt.plot(x)
pandas.DataFrame(x)
numpy.convolve(x,np.ones(4))
numpy.fft.fft(x,n=512)
for i in x:
    print(i)
...

Document

Details can be found here: https://numfi.readthedocs.io/en/latest/?

License

The project is licensed under the MIT license.

benchmark