NumPy integration

It is possible to easily go between numpy and apytypes and therefore also use plotting tools like Matplotlib in an integrated manner.

Consider the example from https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.convolve.html implemented using fixed-point arithmetic

import numpy as np
from scipy import signal
import apytypes as apy
import matplotlib.pyplot as plt

sig = np.repeat([0.0, 1.0, 0.0], 100)
sig_fx = apy.fx(sig, bits=12, int_bits=2)

win = signal.windows.hann(50)
win_fx = apy.fx(win, bits=10, int_bits=2)

filtered = signal.convolve(sig, win, mode="same") / sum(win)
filtered_fx = apy.convolve(sig_fx, win_fx, mode="same") / sum(win_fx)

The results can then be plotted in Matplotlib

fig, (ax_orig, ax_win, ax_filt) = plt.subplots(3, 1, sharex=True)
ax_orig.plot(sig)
ax_orig.plot(sig_fx)
ax_orig.set_title("Original pulse")
ax_orig.margins(0, 0.1)
ax_win.plot(win)
ax_win.plot(win_fx)
ax_win.set_title("Filter impulse response")
ax_win.margins(0, 0.1)
ax_filt.plot(filtered)
ax_filt.plot(filtered_fx)
ax_filt.set_title("Filtered signal")
ax_filt.margins(0, 0.1)
fig.tight_layout()
fig.show()