Signal filtering



In [13]:

    
from scipy.signal import butter
from scipy.signal import lfilter
from sklearn.preprocessing import StandardScaler

import random
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

Filter definition

We define our filter here in which we specify the low cut (0.5Hz), the high cut (30Hz), the frecuency (128Hz) and the order of the filter.

based on: http://scipy-cookbook.readthedocs.io/items/ButterworthBandpass.html



In [15]:

    
def butter_bandpass(lowcut, highcut, fs, order=4):
    nyq = 0.5 * fs
    low = lowcut / nyq
    high = highcut / nyq
    b, a = butter(order, [low, high], btype='band')
    return b, a
b, a = butter_bandpass(0.5, 30.0, 128.0)

Filter application

Here we filter each one of the signal channels and add it to a new dataframe that will be written into a csv file.



In [16]:

    
path = "../../Dataset/Test/EEG_Test_Sorted.csv"
file_name = "Filtered_file.csv"
df = pd.read_csv(path)

labels = ['AF3','F7','F3','FC5','T7','P7','O1','O2','P8','T8','FC6','F4','F8','AF4']
filtered_dataset = pd.DataFrame(columns=['AF3','F7','F3','FC5','T7','P7','O1','O2','P8','T8','FC6','F4','F8','AF4','Class'])

for i in labels:
    temp = lfilter(b, a, df[i])
    filtered_dataset[i] = temp
filtered_dataset['Class'] = df['Class']
filtered_dataset.to_csv(file_name, index=False)