"""
Collection of plot functions.
Each function returns an instance of this module.
"""
import matplotlib.pyplot as plotpy
import numpy
import os
import sys
from dhbw import dasp
[docs]
def exit():
"""
Exit from Python.
"""
sys.exit()
[docs]
def save(path, size=None, dpi=None):
"""
Save current figure to specified image file.
Parameters
----------
path : str
Output image file path.
size : tuple, optional
Output image size (w, h) in inches.
dpi : int, optional
Output image resolution in dots per inch.
"""
assert isinstance(path, str)
assert isinstance(size, (tuple, type(None)))
assert isinstance(dpi, (int, type(None)))
if path.startswith('~'):
path = os.path.expanduser(path)
figure = plotpy.gcf()
origin = figure.get_size_inches()
if size is not None:
figure.set_size_inches(size, forward=True)
if dpi is not None:
figure.savefig(path, dpi=dpi)
else:
figure.savefig(path)
if size is not None:
figure.set_size_inches(origin, forward=True)
return dasp.plot
[docs]
def show(*args, **kwargs):
"""
Display all open figures.
"""
plotpy.show(*args, **kwargs)
return dasp.plot
[docs]
def tight(*args, **kwargs):
"""
Adjust the padding between and around subplots.
"""
plotpy.tight_layout(*args, **kwargs)
return dasp.plot
[docs]
def plot(*args, **kwargs):
"""
Plot y versus x.
"""
plotpy.plot(*args, **kwargs)
return dasp.plot
[docs]
def signal(x, y=None, xlim=None, ylim=1.1):
"""
Plot one or multiple audio signals.
Parameters
----------
x : array or same as y
Timeline array if y is not None.
y : array, tuple or list of arrays, optional
Audio signal(s) to plot, e.g. mono (y) or stereo (y0, y1).
xlim : float, tuple, optional
Time limits.
ylim : float, tuple, optional
Amplitude limits.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(-ylim, +ylim)
if y is None:
assert x is not None
y = x
x = None
if isinstance(y, dict):
axes = []
for i, (k, v) in enumerate(y.items()):
axes += [plotpy.gcf().add_subplot(len(y), 1, i + 1, title=k)]
if x is not None:
plotpy.plot(x, v)
else:
plotpy.plot(v)
lim()
for i in range(1, len(axes)):
axes[i-1].sharex(axes[i])
axes[i-1].sharey(axes[i])
elif isinstance(y, (list, tuple)):
axes = []
for i, v in enumerate(y):
axes += [plotpy.gcf().add_subplot(len(y), 1, i + 1)]
if x is not None:
plotpy.plot(x, v)
else:
plotpy.plot(v)
lim()
for i in range(1, len(axes)):
axes[i-1].sharex(axes[i])
axes[i-1].sharey(axes[i])
else:
if x is not None:
plotpy.plot(x, y)
else:
plotpy.plot(y)
lim()
if x is not None:
plotpy.xlabel('s')
return dasp.plot
[docs]
def spectrogram(x, y, s, t, xlim=None, ylim=None, clim=-120, cmap='inferno', **kwargs):
"""
Plot STFT amplitude of the audio signal.
Parameters
----------
x : array
Timeline array.
y : array
Signal amplitude array.
s : float
STFT step size in seconds.
t : float
STFT time span in seconds.
xlim : float, tuple, optional
Time limits.
ylim : float, tuple, optional
STFT frequency limits in Hz.
clim : float, tuple, optional
STFT amplitude limits in dB.
cmap : str, optional
Plot color map.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(0, ylim)
if clim is not None:
if isinstance(clim, (list, tuple)):
plotpy.clim(clim)
else:
plotpy.clim(clim, 0)
sr = x if numpy.isscalar(x) \
else int(len(x) / numpy.ptp(x)) # 1 / (duration / samples)
fs = int(t * sr) # compute framesize in samples
hs = int(s * sr) # compute hopsize in samples
s = dasp.stft.stft(y, fs, hs)
s = dasp.math.abs(s, db=True)
t = numpy.array([h * hs / sr for h in range(s.shape[0])]) # compute hop timestamps in seconds
f = numpy.fft.rfftfreq(fs, 1 / sr) # compute dft frequencies in hertz
# prefer real coordinates to indices (left, right, bottom, top)
extent = (numpy.min(t), numpy.max(t), numpy.min(f), numpy.max(f))
plotpy.imshow(s.T, aspect='auto', cmap=cmap, extent=extent, interpolation='nearest', origin='lower')
colorbar = plotpy.colorbar()
plotpy.xlabel('s')
plotpy.ylabel('Hz')
colorbar.set_label('dB')
lim()
return dasp.plot
[docs]
def phasogram(x, y, s, t, xlim=None, ylim=None, clim=None, cmap='twilight', **kwargs):
"""
Plot STFT phase of the audio signal.
Parameters
----------
x : array
Timeline array.
y : array
Signal amplitude array.
s : float
STFT step size in seconds.
t : float
STFT time span in seconds.
xlim : float, tuple, optional
Time limits.
ylim : float, tuple, optional
STFT frequency limits in Hz.
clim : float, tuple, optional
STFT phase value limits in dB.
cmap : str, optional
Plot color map.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(0, ylim)
if clim is not None:
if isinstance(clim, (list, tuple)):
plotpy.clim(clim)
else:
plotpy.clim(-clim, +clim)
else:
plotpy.clim(-numpy.pi, +numpy.pi)
sr = x if numpy.isscalar(x) \
else int(len(x) / numpy.ptp(x)) # 1 / (duration / samples)
fs = int(t * sr) # compute framesize in samples
hs = int(s * sr) # compute hopsize in samples
s = dasp.stft.stft(y, fs, hs)
s = dasp.math.arg(s, wrap=True)
t = numpy.array([h * hs / sr for h in range(s.shape[0])]) # compute hop timestamps in seconds
f = numpy.fft.rfftfreq(fs, 1 / sr) # compute dft frequencies in hertz
# prefer real coordinates to indices (left, right, bottom, top)
extent = (numpy.min(t), numpy.max(t), numpy.min(f), numpy.max(f))
plotpy.imshow(s.T, aspect='auto', cmap=cmap, extent=extent, interpolation='nearest', origin='lower')
colorbar = plotpy.colorbar()
plotpy.xlabel('s')
plotpy.ylabel('Hz')
colorbar.set_label('rad')
lim()
return dasp.plot
[docs]
class fft:
[docs]
def abs(x, y, xlim=None, ylim=-120, **kwargs):
"""
Plot DFT amplitude of the audio signal.
Parameters
----------
x : array
Timeline array.
y : array
Signal amplitude array.
xlim : float, tuple, optional
Frequency limits in Hz.
ylim : float, tuple, optional
DFT amplitude limits in dB.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(ylim, 0)
assert x is not None
assert y is not None
if isinstance(y, dict):
for i, (k, v) in enumerate(y.items()):
plotpy.gcf().add_subplot(len(y), 1, i + 1, title=k)
f, a = dasp.fft.abs(x, v, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('dB')
lim()
elif isinstance(y, (list, tuple)):
for i, v in enumerate(y):
plotpy.gcf().add_subplot(len(y), 1, i + 1)
f, a = dasp.fft.abs(x, v, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('dB')
lim()
else:
f, a = dasp.fft.abs(x, y, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('dB')
lim()
plotpy.xlabel('Hz')
return dasp.plot
[docs]
def arg(x, y, xlim=None, ylim=None, **kwargs):
"""
Plot DFT phase of the audio signal.
Parameters
----------
x : array
Timeline array.
y : array
Signal amplitude array.
xlim : float, tuple, optional
Frequency limits in Hz.
ylim : float, tuple, optional
DFT phase value limits.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(+ylim, -ylim)
assert x is not None
assert y is not None
if isinstance(y, dict):
for i, (k, v) in enumerate(y.items()):
plotpy.gcf().add_subplot(len(y), 1, i + 1, title=k)
f, a = dasp.fft.arg(x, v, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('rad')
lim()
elif isinstance(y, (list, tuple)):
for i, v in enumerate(y):
plotpy.gcf().add_subplot(len(y), 1, i + 1)
f, a = dasp.fft.arg(x, v, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('rad')
lim()
else:
f, a = dasp.fft.arg(x, y, **kwargs)
plotpy.plot(f, a)
plotpy.ylabel('rad')
lim()
plotpy.xlabel('Hz')
return dasp.plot
[docs]
class filter:
[docs]
def impulse(b, a, clip=1e3, xlim=None, ylim=None, **kwargs):
"""
Plot impulse response of the transfer function specified by b and a coefficients.
Parameters
----------
b : array
The numerator coefficient array.
a : array
The denominator coefficient array.
clip : float, optional
Clip response amplitudes outside the specified interval.
xlim : float, tuple, optional
Time limits.
ylim : float, tuple, optional
Amplitude limits.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(-ylim, +ylim)
x, y = dasp.filter.impulse(b, a, **kwargs)
if isinstance(clip, (list, tuple)):
y = numpy.clip(y, clip[0], clip[1])
elif clip is not None:
y = numpy.clip(y, -clip, +clip)
plotpy.stem(x, y)
plotpy.xlabel('s')
plotpy.ylabel('')
lim()
return dasp.plot
[docs]
def frequency(b, a, xlim=None, ylim=None, **kwargs):
"""
Plot frequency response of the transfer function specified by b and a coefficients.
Parameters
----------
b : array
The numerator coefficient array.
a : array
The denominator coefficient array.
xlim : float, tuple, optional
Frequency limits in Hz.
ylim : float, tuple, optional
Amplitude value limits in dB.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(ylim, 0)
x, y = dasp.filter.response(b, a, **kwargs)
plotpy.plot(x, dasp.math.abs(y, db=True))
plotpy.xlabel('Hz')
plotpy.ylabel('dB')
lim()
return dasp.plot
[docs]
def phase(b, a, xlim=None, ylim=None, **kwargs):
"""
Plot phase response of the transfer function specified by b and a coefficients.
Parameters
----------
b : array
The numerator coefficient array.
a : array
The denominator coefficient array.
xlim : float, tuple, optional
Frequency limits in Hz.
ylim : float, tuple, optional
Phase value limits in rad.
"""
def lim():
if xlim is not None:
if isinstance(xlim, (list, tuple)):
plotpy.xlim(xlim)
else:
plotpy.xlim(0, xlim)
if ylim is not None:
if isinstance(ylim, (list, tuple)):
plotpy.ylim(ylim)
else:
plotpy.ylim(-ylim, +ylim)
else:
plotpy.ylim(-numpy.pi, +numpy.pi)
x, y = dasp.filter.response(b, a, **kwargs)
plotpy.plot(x, dasp.math.arg(y, wrap=True))
plotpy.xlabel('Hz')
plotpy.ylabel('rad')
lim()
return dasp.plot
[docs]
def poleszeros(b, a):
"""
Plot poles and zeros in Z plane for the transfer function specified by b and a coefficients.
Parameters
----------
b : array
The numerator coefficient array.
a : array
The denominator coefficient array.
"""
# https://gist.github.com/endolith/4625838
p = dasp.filter.poles(b, a)
z = dasp.filter.zeros(b, a)
c = plotpy.Circle((0, 0), radius=1, linestyle=':', linewidth=1, color='black', fill=False, alpha=0.3)
r = numpy.amax(numpy.concatenate((numpy.abs(p), numpy.abs(z), [1]))) + 0.1
axis = plotpy.gca()
axis.add_patch(c)
axis.axvline(0, linestyle=':', linewidth=1, color='black', alpha=0.3)
axis.axhline(0, linestyle=':', linewidth=1, color='black', alpha=0.3)
if len(p):
axis.plot(p.real, p.imag, 'x', markersize=10, color='red', alpha=0.7, label='Pole')
if len(z):
axis.plot(z.real, z.imag, 'o', markersize=10, color='none', markeredgecolor='blue', alpha=0.7, label='Zero')
axis.legend()
axis.set_xlabel('Re(z)')
axis.set_ylabel('Im(z)')
axis.axis('scaled')
axis.axis([-r, +r, -r, +r])
return dasp.plot