Add files via upload

use np.convolve to apply a moving average filter

moving_average_filter.py

+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Created on Tue Oct 10 16:49:23 2017
+
+@author: Neo
+
+Apply a moving average filter to an array.
+
+"""
+
+import numpy as np
+import matplotlib.pyplot as plt
+
+
+# ------------------------------  MAIN BODY  -----------------------------
+# Generate a Gaussioan white nosie.
+N = 500
+mu, sigma = 0, 1  # mean and standard deviation
+w = np.random.normal(mu, sigma, N)
+# Use a moving average filter v[t] = (w[t-1] + w[t] + w[t+1]) / 3.0
+# 1) a normal way
+v = np.empty(N)
+# For head
+v[0] = (w[0] + w[1]) / 3.0
+# For tail
+v[-1] = (w[-2] + w[-1]) / 3.0
+# For else
+v[1:-1] = (w[:-2] + w[1:-1] + w[2:]) / 3.0
+# 2) use np.convolve() function
+n = 3
+window = np.ones(n) * 1.0 / n
+v1 = np.convolve(window, w, mode='same')
+# Plot the result.
+plt.plot(v1 - v)
+plt.xlim([0, N])
+plt.legend()
+plt.show()
+print('Done!')
+# ------------------------------ END -----------------------------------