Filter Time Series data using NumPy and SciPy Signal processing

Last update on April 30 2025 12:55:53 (UTC/GMT +8 hours)

5. Signal Processing on Time Series Data

Write a NumPy program to create a time series dataset and apply SciPy's signal processing functions to filter the data.

Sample Solution:

Python Code:

# Import necessary libraries
import numpy as np
from scipy import signal
import matplotlib.pyplot as plt

# Create a time series dataset using NumPy
np.random.seed(0)  # For reproducibility
time = np.linspace(0, 1, 500, endpoint=False)
data = np.sin(2 * np.pi * 7 * time) + np.random.randn(500) * 0.5

# Design a Butterworth filter using SciPy
b, a = signal.butter(4, 0.1)

# Apply the filter to the time series data
filtered_data = signal.filtfilt(b, a, data)

# Plot the original and filtered data
plt.plot(time, data, label='Original Data')
plt.plot(time, filtered_data, label='Filtered Data', linewidth=2)
plt.legend()
plt.show()

Output:

Explanation:

Import necessary libraries:

Import NumPy, SciPy's signal module, and Matplotlib for plotting.

Create a time series dataset using NumPy:

Generate time series data consisting of a sine wave with added noise.

Design a Butterworth filter using SciPy:

Use SciPy's butter function to design a low-pass Butterworth filter.

Apply the filter to the time series data:

Use SciPy's "filtfilt()" function to apply the filter, ensuring zero phase distortion.

Plot the original and filtered data:

Use Matplotlib to visualize both the original noisy data and the filtered data.

For more Practice: Solve these Related Problems:

Write a Numpy program to generate a synthetic time series and apply a low-pass filter using SciPy's signal module.
Write a Numpy program to simulate a noisy signal and remove noise using a band-stop filter from SciPy's signal processing tools.
Write a Numpy program to compute and plot the frequency response of a filter designed with SciPy's signal module.
Write a Numpy program to design and apply a digital filter on a time series, then compare the filtered output with the original signal.

Go to:

Previous: Interpolate data points using NumPy and SciPy's Interpolate module.
Next: Fit a curve to sample data using NumPy and SciPy's curve_fit.

Python-Numpy Code Editor:

Have another way to solve this solution? Contribute your code (and comments) through Disqus.