convert_to_spectrogram.py

## Importing all Libraries/Dependencies 

import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
from scipy.io import wavfile
import os
import csv
from sklearn.preprocessing import normalize
import numpy as np

## 	The folder structure should be set up maually before running
## 	this script according to the project notebook 'CardiacML.ipynb'

'''
	 Convert the recordings to spectrogram and save it to 
	 seperate classes/folders

	params@ :
		wav_file - recording file 
		class_label - label for the recording i.e 'normal' or 'abnormal'  
'''
def graph_spectrogram(wav_file, class_label):
    rate, data = get_wav_info(wav_file)
    nfft = 256  # Length of the windowing segments
    fs = 256    # Sampling frequency
    pxx, freqs, bins, im = plt.specgram(data, nfft,fs)
    plt.axis('off')
    if (class_label == 1):
    	plt.savefig('d/abnormal/' + str(wav_file[23:28]) + '.png',
                dpi=100, # Dots per inch
                frameon='false',
                aspect='normal',
                bbox_inches='tight',
                pad_inches=0) # Spectrogram saved as a .png
    else:
    	plt.savefig('d/normal/' + str(wav_file[23:28]) + '.png',
                dpi=100, # Dots per inch
                frameon='false',
                aspect='normal',
                bbox_inches='tight',
                pad_inches=0)
    		


'''
	Returns the sampling rate and data of the recording 
	params@ :
		wav_file - recording file 
'''   		
def get_wav_info(wav_file):
    rate, data = wavfile.read(wav_file)
    return rate, data


'''
	Returns the class label i.e 'normal' or 'abnormal' 

	params@ :
		label_file_name - label (.hea) file for the recording  
'''
def parse_class_label(label_file_name):

    with open(label_file_name, 'r') as fin:
        header = fin.readlines()

    comments = [line for line in header if line.startswith("#")]
    if not len(comments) == 1:
        raise InvalidHeaderFileException("Invalid label file %s" % label_file_name)

    class_label = str(comments[0]).lstrip("#").rstrip("\r").strip().lower()

    if not class_label in class_name_to_id.keys():
        raise InvalidHeaderFileException("Invalid class label %s" % class_label)

    return class_label 



if __name__ == "__main__":

	class_name_to_id = {"normal": 0, "abnormal": 1}
	nclasses = len(class_name_to_id.keys())
	path = '../training/'
	wav_file_names = []
	class_labels = []
	samples = 0
	## processing all the recordings in 'training' folder
	for root, dirs, files in os.walk(path):
	    for file in files:
	        if file.endswith('.wav'):
	            base_file_name = file.rstrip(".wav")
	            label_file_name = os.path.join(root, base_file_name + ".hea")
	            class_label = parse_class_label(label_file_name)
	            class_labels.append(class_name_to_id[class_label])
	            wav_file_names.append(os.path.join(root, file))
	            samples +=1


	## After saving all file name to lists, convering to spectrograms
	for i,j in zip(wav_file_names,class_labels):
	    graph_spectrogram(i,j)

	## Logging the number of samples and file names   
	print (len(class_labels), len(wav_file_names))
	for i,j in zip(wav_file_names[:10], class_labels[:10]):
		print (i[23:28], j)


	## saving the labels to a csv file
	with open('labels.csv', 'w') as f:
	    writer = csv.writer(f)
	    for val in class_labels:
	        writer.writerow([val])