示例#1
0
def test_compare_outputs():
    """Compare the python output with the matlab output.
    They should be identical."""
    channel_mat = 'matlab/codes/channel/channel.mat'

    mat_dict = scipy.io.loadmat(channel_mat)

    UstarX1 = mat_dict['UstarX1']
    S = mat_dict['S']
    V = mat_dict['V']

    # Sparsity-promoting parameter gamma
    # Lower and upper bounds relevant for this flow type
    gamma_grd = 20
    gammaval = np.logspace(np.log10(0.15), np.log10(160), gamma_grd)

    Fdmd, Edmd, Ydmd, xdmd, py_answer = sparse_dmd.run_dmdsp(UstarX1,
                                                            S,
                                                            V,
                                                            gammaval)
    # convert class attributes to dict
    py_answer = py_answer.__dict__

    ## matlab output created by running matlab:
    ## [Fdmd, Edmd, Ydmd, xdmd, answer] = run_dmdsp;
    ## output = struct('Fdmd', Fdmd, 'Edmd', Edmd, 'Ydmd', Ydmd, 'xdmd', xdmd);
    ## save('tests/answer.mat', '-struct', 'answer')
    ## save('tests/output.mat', '-struct', 'output')
    ## using channel=1 and 20 gridpoints for gamma

    # using two different .mat because scipy doesn't seem to like
    # nested structs
    mat_answer = scipy.io.loadmat('tests/answer.mat')
    mat_output = scipy.io.loadmat('tests/output.mat')

    for k in py_answer:
        nt.assert_array_almost_equal(py_answer[k].squeeze(),
                                     mat_answer[k].squeeze(),
                                     decimal=5)

    py_output = {'Fdmd': Fdmd, 'Edmd': Edmd, 'Ydmd': Ydmd, 'xdmd': xdmd}

    for k in py_output:
        nt.assert_array_almost_equal(py_output[k].squeeze(),
                                     mat_output[k].squeeze(),
                                     decimal=5)
示例#2
0
def test_compare_outputs():
    """Compare the python output with the matlab output.
    They should be identical."""
    channel_mat = 'matlab/codes/channel/channel.mat'

    mat_dict = scipy.io.loadmat(channel_mat)

    UstarX1 = mat_dict['UstarX1']
    S = mat_dict['S']
    V = mat_dict['V']

    # Sparsity-promoting parameter gamma
    # Lower and upper bounds relevant for this flow type
    gamma_grd = 20
    gammaval = np.logspace(np.log10(0.15), np.log10(160), gamma_grd)

    Fdmd, Edmd, Ydmd, xdmd, py_answer = sparse_dmd.run_dmdsp(
        UstarX1, S, V, gammaval)
    # convert class attributes to dict
    py_answer = py_answer.__dict__

    ## matlab output created by running matlab:
    ## [Fdmd, Edmd, Ydmd, xdmd, answer] = run_dmdsp;
    ## output = struct('Fdmd', Fdmd, 'Edmd', Edmd, 'Ydmd', Ydmd, 'xdmd', xdmd);
    ## save('tests/answer.mat', '-struct', 'answer')
    ## save('tests/output.mat', '-struct', 'output')
    ## using channel=1 and 20 gridpoints for gamma

    # using two different .mat because scipy doesn't seem to like
    # nested structs
    mat_answer = scipy.io.loadmat('tests/answer.mat')
    mat_output = scipy.io.loadmat('tests/output.mat')

    for k in py_answer:
        nt.assert_array_almost_equal(py_answer[k].squeeze(),
                                     mat_answer[k].squeeze(),
                                     decimal=5)

    py_output = {'Fdmd': Fdmd, 'Edmd': Edmd, 'Ydmd': Ydmd, 'xdmd': xdmd}

    for k in py_output:
        nt.assert_array_almost_equal(py_output[k].squeeze(),
                                     mat_output[k].squeeze(),
                                     decimal=5)
示例#3
0
def create_answer(n=200):
    answer = sparse_dmd.run_dmdsp(gamma_grd=n)
    scipy.io.savemat('tests/answer.mat', answer)
示例#4
0
def create_answer(n=200):
    answer = sparse_dmd.run_dmdsp(gamma_grd=n)
    scipy.io.savemat('tests/answer.mat', answer)
示例#5
0
import time

import numpy as np
import scipy.io

from sparse_dmd import run_dmdsp

channel_mat = 'matlab/codes/channel/channel.mat'

mat_dict = scipy.io.loadmat(channel_mat)

UstarX1 = mat_dict['UstarX1']
S = mat_dict['S']
V = mat_dict['V']

# Sparsity-promoting parameter gamma
# Lower and upper bounds relevant for this flow type
gamma_grd = 200
gammaval = np.logspace(np.log10(0.15), np.log10(160), gamma_grd)

tic = time.time()
Fdmd, Edmd, Ydmd, xdmd, py_answer = run_dmdsp(UstarX1, S, V, gammaval)
toc = time.time()
print "time elapsed: ", toc - tic