#!/usr/bin/env python
#
# pyLOM - Python Low Order Modeling.
#
# DMD plotting utilities.
#
# Last rev: 27/10/2021
from __future__ import print_function, division
import numpy as np
import matplotlib.pyplot as plt
from .utils import extract_modes
from ..utils import gpu_to_cpu
from ..utils.plots import plotResidual, plotFieldStruct2D, plotSnapshot, plotLayout
[docs]
def plotMode(Phi, omega, mesh, dset, ivar, pointData=True, modes=np.array([1],np.int32),**kwargs):
'''
Plot the real and imaginary parts of a mode
'''
Phi, omega = gpu_to_cpu(Phi), gpu_to_cpu(omega)
# Extract the modes to be plotted
npoints = mesh.size(pointData)
Phi_real = extract_modes(Phi,ivar,npoints,real=True,modes=modes)
Phi_imag = extract_modes(Phi,ivar,npoints,real=False,modes=modes)
# Add to the dataset
dset.add_field('PHI_REAL',len(modes),Phi_real)
dset.add_field('PHI_IMAG',len(modes),Phi_imag)
# Loop over the modes
screenshot = kwargs.pop('screenshot',None)
off_screen = kwargs.pop('off_screen',False)
for imode, mode in enumerate(modes):
if screenshot is not None: kwargs['screenshot'] = screenshot % imode
plotLayout(mesh,dset,2,1,mode-1,vars=['PHI_REAL','PHI_IMAG'],title='Mode %d St = %.3f' % (mode-1, np.abs(omega[mode-1])/(2*np.pi)),off_screen=off_screen,**kwargs)
# Remove from dataset
dset.delete('PHI_REAL')
dset.delete('PHI_IMAG')
[docs]
def ritzSpectrum(real, imag, fig = None, ax = None, cmap = None):
'''
Given the real and imaginary part of the eigenvalues, plot the Ritz Spectrum together with the unit circle
'''
real, imag = gpu_to_cpu(real), gpu_to_cpu(imag)
if fig is None:
fig = plt.figure(figsize=(8,6),dpi=100)
if ax is None:
ax = fig.add_subplot(1,1,1)
if cmap is None:
cmap = 'r'
theta = np.array(range(101))*2*np.pi/100
ax.plot(np.cos(theta), np.sin(theta), c = 'k')
ax.scatter(real, imag, c = cmap)
ax.axis('equal')
ax.set(xlabel = r'$\mu_{Re}$', ylabel = r'$\mu_{Imag}$', title = 'Ritz spectrum')
return fig, ax
[docs]
def amplitudeFrequency(omega, amplitude, fig = None, ax = None, cmap = None, mark = None, norm = False):
'''
Given the frequency and the amplitude of the DMD modes, plot the amplitude against the Strouhal number
'''
omega, amplitude = gpu_to_cpu(omega), gpu_to_cpu(amplitude)
if fig is None:
fig = plt.figure(figsize=(8,6),dpi=100)
if ax is None:
ax = fig.add_subplot(1,1,1)
if cmap is None:
cmap = 'r'
if mark is None:
mark = 'X'
if norm is True:
amplitude = np.abs(amplitude)/np.max(np.abs(amplitude))
else:
amplitude = np.abs(amplitude)
ax.scatter(omega/(2*np.pi), amplitude, c = cmap, marker = mark)
ax.set(xlabel = 'St', ylabel = 'Amplitude', title = 'Amplitude vs Frequency of the DMD Modes')
ax.set_yscale('log')
return fig, ax
[docs]
def dampingFrequency(omega, delta, fig = None, ax = None, cmap = None, mark = None):
'''
Given the frequency and the damping ratio of the DMD modes, plot the amplitude against the Strouhal number
'''
omega, delta = gpu_to_cpu(omega), gpu_to_cpu(delta)
if fig is None:
fig = plt.figure(figsize=(8,6),dpi=100)
if ax is None:
ax = fig.add_subplot(1,1,1)
if cmap is None:
cmap = 'r'
if mark is None:
mark = 'X'
ax.scatter(omega/(2*np.pi), np.abs(delta), c = cmap, marker = mark)
ax.set(xlabel = 'St', ylabel = 'Damping Ratio', title = 'Damping ratio vs Frequency of the DMD Modes')
ax.set_yscale('log')
return fig, ax
