def optimize_components(X, feature_names, label, abbrev):
# model selection (optimal number of components)
# from https://scikit-learn.org/stable/auto_examples/decomposition/plot_pca_vs_fa_model_selection.html
# prepare explained variance scorer
def get_score(model, data, scorer=explained_variance_score):
prediction = model.inverse_transform(model.transform(data))
return scorer(data, prediction)
# choose number of components by explained variance
n_components = np.arange(1, len(feature_names) + 1)
nmf = NMF(random_state=SEED)
nmf_scores = []
for n in n_components:
nmf.n_components = n
nmf.fit(X)
nmf_scores.append(get_score(nmf, X))
nmf_scores = np.array(nmf_scores)
n_components_nmf = n_components[np.argmax(nmf_scores >= 0.95)]
print(label + ": best n_components by explained variance > 0.95 = %d" %
int(n_components_nmf))
# create plot
plt.figure()
plt.plot(n_components,
nmf_scores,
'b',
label='explained variance by num/components')
plt.axvline(n_components_nmf,
color='b',
label='chosen number of components: %d' % n_components_nmf,
linestyle='--')
# format plot
ax = plt.gca()
ax.xaxis.set_major_locator(MaxNLocator(integer=True))
plt.xlabel('number of components')
plt.ylabel('explained variance')
plt.legend(loc='lower right')
plt.title(label + ": NMF model selection")
plt.savefig(path.join(PLOT_DIR, abbrev + "_nmf_components.png"),
bbox_inches='tight')
plt.show()
plt.close()
return n_components_nmf
def compute_scores(X):
pca = PCA(svd_solver='auto')
kpca = KernelPCA(fit_inverse_transform=True)
ica = FastICA()
nmf = NMF(init='nndsvda')
pca_scores, ica_scores, nmf_scores, kpca_scores = [], [], [], []
for n in n_components:
pca.n_components = n
ica.n_components = n
nmf.n_components = n
kpca.n_components = n
print(n)
Xpca = pca.inverse_transform(pca.fit_transform(Xs))
pca_scores.append(explained_variance_score(Xs, Xpca))
Xica = ica.inverse_transform(ica.fit_transform(Xs))
ica_scores.append(explained_variance_score(Xs, Xica))
Xkpca = kpca.inverse_transform(kpca.fit_transform(Xs))
kpca_scores.append(explained_variance_score(Xs, Xkpca))
Xnmf = nmf.inverse_transform(nmf.fit_transform(X))
nmf_scores.append(explained_variance_score(X, Xnmf))
return pca_scores, ica_scores, nmf_scores, kpca_scores
