-
Notifications
You must be signed in to change notification settings - Fork 0
/
eda.py
162 lines (129 loc) · 4.62 KB
/
eda.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon Mar 23 10:06:33 2020
@author: tsuyogbasnet
"""
import os
from tqdm import tqdm
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from scipy.io import wavfile
from python_speech_features import mfcc, logfbank
import librosa
#calculate fft @params signal and rate
def calculate_fft(signal, rate):
signal_length = len(signal)
frequency = np.fft.rfftfreq(signal_length, d = 1/rate)
#mean normalization of length of signal
magnitude = abs(np.fft.rfft(signal)/signal_length)
return (magnitude, frequency)
#calculate low and hig frequency because we have dead spots in signal
#we do this to get atleats a threshold value of amps
def envelope(signal, rate, threshold):
mask = []
signal = pd.Series(signal).apply(np.abs)
signal_mean = signal.rolling(window=int(rate/10), min_periods = 1, center = True).mean()
for mean in signal_mean:
if mean > threshold:
mask.append(True)
else:
mask.append(False)
return mask
def plot_signals(signals):
fig, axes = plt.subplots(nrows=2, ncols=5, sharex=False,
sharey=True, figsize=(20,5))
fig.suptitle('Time Series', size=16)
i = 0
for x in range(2):
for y in range(5):
axes[x,y].set_title(list(signals.keys())[i])
axes[x,y].plot(list(signals.values())[i])
axes[x,y].get_xaxis().set_visible(False)
axes[x,y].get_yaxis().set_visible(False)
i += 1
def plot_fft(fft):
fig, axes = plt.subplots(nrows=2, ncols=5, sharex=False,
sharey=True, figsize=(20,5))
fig.suptitle('Fourier Transforms', size=16)
i = 0
for x in range(2):
for y in range(5):
data = list(fft.values())[i]
Y, freq = data[0], data[1]
axes[x,y].set_title(list(fft.keys())[i])
axes[x,y].plot(freq, Y)
axes[x,y].get_xaxis().set_visible(False)
axes[x,y].get_yaxis().set_visible(False)
i += 1
def plot_fbank(fbank):
fig, axes = plt.subplots(nrows=2, ncols=5, sharex=False,
sharey=True, figsize=(20,5))
fig.suptitle('Filter Bank Coefficients', size=16)
i = 0
for x in range(2):
for y in range(5):
axes[x,y].set_title(list(fbank.keys())[i])
axes[x,y].imshow(list(fbank.values())[i],
cmap='hot', interpolation='nearest')
axes[x,y].get_xaxis().set_visible(False)
axes[x,y].get_yaxis().set_visible(False)
i += 1
def plot_mfccs(mfccs):
fig, axes = plt.subplots(nrows=2, ncols=5, sharex=False,
sharey=True, figsize=(20,5))
fig.suptitle('Mel Frequency Cepstrum Coefficients', size=16)
i = 0
for x in range(2):
for y in range(5):
axes[x,y].set_title(list(mfccs.keys())[i])
axes[x,y].imshow(list(mfccs.values())[i],
cmap='hot', interpolation='nearest')
axes[x,y].get_xaxis().set_visible(False)
axes[x,y].get_yaxis().set_visible(False)
i += 1
data_frame = pd.read_csv('instruments.csv')
data_frame.set_index('fname', inplace=True)
for f in data_frame.index:
rate, signal = wavfile.read('sounds/'+f)
data_frame.at[f, 'length'] = signal.shape[0]/rate
classes = list(np.unique(data_frame.label))
class_dist = data_frame.groupby(['label'])['length'].mean()
#plotting the data
fig, ax = plt.subplots()
ax.set_title("class distribution", y=1.08)
ax.pie(class_dist, labels=class_dist.index, autopct='%1.1f%%', shadow=False, startangle=90)
ax.axis('equal')
#plt.show()
data_frame.reset_index(inplace=True)
##datapoints
signals = {}
fft = {}
fbank = {}
mfccs = {}
for c in classes:
wav_file = data_frame[data_frame.label == c].iloc[0,0]
signal, rate = librosa.load('sounds/'+wav_file, sr=44100)
mask = envelope(signal, rate, 0.0005)
signal = signal[mask]
signals[c] = signal
fft[c] = calculate_fft(signal, rate)
bank = logfbank(signal[:rate], rate, nfilt=26, nfft=1103).T
fbank[c] = bank
mel = mfcc(signal[:rate], rate, numcep=13, nfilt = 26, nfft = 1103).T
mfccs[c] = mel
plot_signals(signals)
plt.show()
plot_fft(fft)
plt.show()
plot_fbank(fbank)
plt.show()
plot_mfccs(mfccs)
plt.show()
#stroing cleaned data for modelling
if len(os.listdir('cleanfiles')) == 0:
for f in tqdm(data_frame.fname):
signal, rate = librosa.load('sounds/'+f, sr = 16000)
mask = envelope(signal, rate, 0.0005)
wavfile.write(filename='cleanfiles/'+f, rate = rate, data= signal[mask])