def test_normalize_transform(self):
"""Is the transformed batch file different from the original one?"""
inputs = {
'biom_file': None,
'cluster': ['Affinity'],
'nclust': ['4'],
'otu_meta': None,
'otu_table': ['otu_bananas.txt'],
'prefix': None,
'sample_data': None,
'split': 'Rocket Science',
'tax_table': ['tax_bananas.txt'],
'name': ['test'],
'fp': os.path.dirname(massoc.__file__)[:-7].replace('\\', '/')
}
batch = Batch(testbiom, inputs)
clrbatch = batch.normalize_transform(mode="clr")
self.assertFalse(batch.otu['test'] == clrbatch.otu['test'])
def generate_cluster_figures(self):
"""Generates figures for diagnostics canvas."""
from massoc.scripts.batch import Batch
from sklearn.cluster import KMeans, DBSCAN, SpectralClustering, AffinityPropagation
from sklearn.mixture import GaussianMixture
from sklearn.metrics import silhouette_score
from sklearn.decomposition import PCA
nums = list(range(2, 5))
try:
file = self.file_list.GetSelection()
file = self.file_list.GetString(file)
x = 'init'
biomfile = {x: biom.load_table(file)}
algo = self.cluster_choice.GetSelection()
algo = self.cluster_choice.GetString(algo)
inputs = {'biom_file': [file],
'cluster': [algo]}
normbatch = Batch(biomfile, inputs)
normbatch = normbatch.normalize_transform(mode='clr')
norm_table = normbatch.otu[x]
topscore = 0
bestcluster = [1] * len(norm_table.ids())
data = csr_matrix.todense(norm_table.matrix_data)
data = np.matrix.transpose(data)
data = PCA(n_components=2).fit_transform(data)
randomclust = np.random.randint(2, size=len(data))
sh_score = [silhouette_score(data, randomclust)]
# K-means clustering, tests 2-4 clusters
if inputs['cluster'][0] == 'K-means':
for i in nums:
clusters = KMeans(i).fit_predict(data)
silhouette_avg = silhouette_score(data, clusters)
sh_score.append(silhouette_avg)
topscore = int(np.argmax(sh_score) + 1)
bestcluster = KMeans(topscore).fit_predict(data)
# DBSCAN clustering, automatically finds optimal cluster size
if inputs['cluster'][0] == 'DBSCAN':
bestcluster = DBSCAN().fit_predict(data)
topscore = len(set(bestcluster)) - (1 if -1 in bestcluster else 0)
# Gaussian Mixture Model (gmm) probability distribution
if inputs['cluster'][0] == 'Gaussian':
for i in nums:
fit = GaussianMixture(i).fit(data)
clusters = fit.predict(data)
silhouette_avg = silhouette_score(data, clusters)
sh_score.append(silhouette_avg)
topscore = int(np.argmax(sh_score) + 1)
bestfit = GaussianMixture(topscore).fit(data)
bestcluster = bestfit.predict(data)
# Spectral Clustering
if inputs['cluster'][0] == 'Spectral':
for i in nums:
clusters = SpectralClustering(i).fit_predict(data)
silhouette_avg = silhouette_score(data, clusters)
sh_score.append(silhouette_avg)
topscore = int(np.argmax(sh_score) + 1)
bestcluster = SpectralClustering(topscore).fit_predict(data)
# Affinity Propagation clustering
if inputs['cluster'] == 'Affinity':
bestcluster = AffinityPropagation().fit_predict(data)
topscore = len(set(bestcluster)) - (1 if -1 in bestcluster else 0)
if max(sh_score) < 0.25:
raise ValueError("Silhouette score too low: please try a different algorithm. "
"Your data may not be suitable for clustering.")
for i in range(topscore):
mask, = np.where(bestcluster == i)
for j in mask:
norm_table._sample_metadata[j]['cluster'] = inputs['cluster'][0] + '_' + str(i)
x, y = zip(*data)
self.prev.scatter(x, y, bestcluster)
self.canvas1.draw()
except Exception:
logger.error("Failed to generate figures. ", exc_info=True)