S3SignalUtils.py

#!/usr/bin/env python3
"""Utils function related to signals for S3"""

from typing import Tuple

import scipy.signal as sig
from scipy.signal import butter, freqz, lfilter

from dependency import np, wavutils


def sigin(wavname: str) -> Tuple[int, np.ndarray]:
    """Functions that reads wave file and return sample rate and signal as np.array"""
    return wavutils.read(wavname, mmap=False)


def sawtooth(fs: float, Ns: int, Ss: int) -> np.ndarray:
    """Returns a Sawtooth wave of Sample rate Ss with Ns number of samples and Sample Frequency Fs"""
    Ss = np.linspace(0, 1, Ss)
    return sig.sawtooth(2 * np.pi * fs * Ss)[0:Ns]


def triangle(fs: float, Ns: int, Ss: int) -> np.ndarray:
    """Returns a Triangle wave of Sample rate Ss with Ns number of samples and Sample Frequency Fs"""
    Ss = np.linspace(0, 1, Ss)
    return sig.sawtooth(2 * np.pi * fs * Ss, 0.5)[0:Ns]


def sin(fs: float, Ns: int, Ss: int) -> np.ndarray:
    """Returns a Sine wave of Sample rate Ss with Ns number of samples and Sample Frequency Fs"""
    t = np.arange(Ns)
    omega = 2 * np.pi * fs / Ss
    return np.sin(omega * t)


def cos(fs: float, Ns: int, Ss: int) -> np.ndarray:
    """Returns a Cosine wave of Sample rate Ss with Ns number of samples and Sample Frequency Fs"""
    t = np.arange(Ns)
    omega = 2 * np.pi * fs / Ss
    return np.cos(omega * t)


# Check tests3 in signals-systems to write these functions


def filt_lp(sig: np.ndarray, Ss: int, Cfs: int, Cfs1: None,
            order=5) -> np.ndarray:
    """return a filtered signal; low pass filter"""
    nyq = 0.5 * Ss
    normal_cutoff = Cfs / nyq
    b, a = butter(order, normal_cutoff, btype='low', analog=False)
    return lfilter(b, a, sig)


def filt_hp(sig: np.ndarray, Ss: int, Cfs: int, Cfs1: None,
            order=5) -> np.ndarray:
    """return a filtered signal; high pass filter"""
    nyq = 0.5 * Ss
    normal_cutoff = Cfs / nyq
    b, a = butter(order, normal_cutoff, btype='high', analog=False)
    return lfilter(b, a, sig)


def filt_bp(sig: np.ndarray, Ss: int, Cfs0: int, Cfs1: None,
            order=5) -> np.ndarray:
    """return a filtered signal; band pass filter"""
    nyq = 0.5 * Ss
    normal_cutoff1 = Cfs0 / nyq
    normal_cutoff2 = Cfs1 / nyq
    b, a = butter(order, (normal_cutoff1, normal_cutoff2),
                  btype='band',
                  analog=False)
    return lfilter(b, a, sig)