xs = fitted[:, 0]
ys = fitted[:, 1]
zs = fitted[:, 2]
utils.plot_3d_mod(xs,
ys,
zs,
title='wf{} {}'.format(wf, scid),
s=5,
edge='navy',
xlabel='PC1',
ylabel='PC2',
zlabel='PC3',
add_info='PAT{} ({})'.format(wf, scid))
if input('Satisfied? (y|N)\n') != 'y':
raise Exception('Not satisified :(')
# =============================================================================
# Gaussian Mixture Models and, if needed, additional clustering
n_components = int(input('Type the number of components to split into: '))
fitted_copy = fitted
cluster_means, clusters, ns, ks, clusters_list = utils.gaussian_mixture_pca_projections(
fitted,
waves,
n_components,
npat=None,
scid=scid,
wave_type=wave_type,
n_init=5)
zlabel = 'PC3',
add_info = 'PAT{} ({})'.format(patient_nr, scid))
if input('Satisfied? (y|N)\n') != 'y':
raise Exception('Not satisified :(')
"""
# =============================================================================
# Gaussian Mixture Models and, if needed, additional clustering
n_components = int(input('Type the number of components to split into: '))
fitted_copy = fitted
cluster_means, clusters, ns, ks, clusters_list = utils.gaussian_mixture_pca_projections(
fitted,
waves,
n_components,
npat=patient_nr,
scid=scid,
wave_type=wave_type,
n_init=1,
random_state=120000)
# =============================================================================
# Plot random ABP waves and their ICP counterparts
"""
utils.plot_random_wave_samples(waves, wave_type='abp', seg_type=scid, rand_list=clusters_list[-1])
icp_waves = utils.acces_waves_by_type(None, stacked_waves, 'icp')
utils.plot_random_wave_samples(icp_waves, wave_type='icp', seg_type=scid, rand_list=clusters_list[-1])
"""
# =============================================================================
# Fit-Transforming PCA on acquired waves:
print('\n\nAnalysis of corresponding counterparts:')
wave_type = 'icp'
add_info = '({})'.format(scid))
if input('\nSatisfied? (y|N)\n') != 'y':
raise Exception('Not satisified :(')
"""
# =============================================================================
# Gaussian Mixture Models OR HDBSCAN and, if needed, additional clustering *
decision = input('dbscan or gmm: ')
if decision == 'gmm':
n_components = int(input('Type the number of components to split into: '))
cluster_means, clusters, ns, ks, clusters_list = utils.gaussian_mixture_pca_projections(
fitted,
waves,
n_components,
npat=None,
scid=scid,
wave_type='icp',
n_init=1,
random_state=3)
print('That was it.')
# ______________________________________________________________________________
elif decision == 'dbscan':
labels, sub_colors, n_clusters = utils.hdbscan_(fitted)
xs = fitted[:, 0]
ys = fitted[:, 1]
zs = fitted[:, 2]
N = len(np.unique(sub_colors))
labels += 1
#print(waves[2 == np.where(wf_types == 1)].shape)
utils.plot_3d(xs,
ys,
zs,
title='NFPAT{} {} PCA projection'.format(patient_nr, scid),
xlabel='PC1',
ylabel='PC2',
zlabel='PC3',
s=1)
# Gaussian Mixture Models and, if needed, additional clustering
n_components = int(input('Type the number of components to split into: '))
cluster_means, clusters, ns, ks = utils.gaussian_mixture_pca_projections(
fitted,
concatenated_waves,
n_components,
patient_nr,
scid,
with_mahal=False)
clusters = np.asarray(clusters)
if input('Continue with cleaning and saving? ') != 'yes':
import sys
sys.exit()
else:
# Cleaning (only if additional clustering was needed):
if ns is not None:
print('\nCleaning...')
x = cluster_means.shape[0]
scid = 'nsc'
utils.plot_3d(xs,
ys,
zs,
title='PCA projection'.format(scid),
xlabel='PC1',
ylabel='PC2',
zlabel='PC3',
s=1)
# Gaussian Mixture Models and, if needed, additional clustering
n_components = int(input('Type the number of components to split into: '))
cluster_means, clusters, ns, ks = utils.gaussian_mixture_pca_projections(
fitted, concatenated_waves, n_components)
"""
clusters = np.asarray(clusters)
# Cleaning (only if additional clustering was needed):
if ns is not None:
print('\nCleaning...')
x = cluster_means.shape[0]
# Means:
cluster_means = np.delete(cluster_means, ns, axis=0)
# Clusters:
clusters = np.delete(clusters, ns, axis=0)
assert cluster_means.shape[0] == clusters.shape[0]
print('{} cluster(s) removed!'.format(x - cluster_means.shape[0]))