SciPy — Signal Processing

Filter design

from scipy.signal import butter, filtfilt
import numpy as np

# Design lowpass filter
b, a = butter(4, 0.1, btype='low')

# Apply filter
signal = np.random.randn(1000)
filtered = filtfilt(b, a, signal)

FIR filter

from scipy.signal import firwin, lfilter

# Design FIR filter
numtaps = 51
cutoff = 0.2
taps = firwin(numtaps, cutoff)

# Apply filter
filtered = lfilter(taps, 1.0, signal)

Frequency response

from scipy.signal import freqz

w, h = freqz(b, a)

Spectrogram

from scipy.signal import spectrogram
import numpy as np

fs = 1000  # Sample frequency
t = np.linspace(0, 1, fs)
signal = np.sin(2 * np.pi * 100 * t)

f, t_spec, Sxx = spectrogram(signal, fs)

Peak finding

from scipy.signal import find_peaks

signal = np.sin(np.linspace(0, 10, 1000))
peaks, _ = find_peaks(signal, height=0.5)
print(f"Found {len(peaks)} peaks")

Convolution

from scipy.signal import convolve

a = np.array([1, 2, 3])
b = np.array([4, 5, 6])
c = convolve(a, b)
print(f"Convolution: {c}")

Applications

• Audio filtering
• Noise removal
• Feature extraction
• Communication systems

Mini Practice

1. Design a lowpass filter
2. Filter a noisy signal
3. Create a spectrogram
4. Find peaks in data

Up Next

Continue with Statistics - Statistical functions.