def community_detection(self, args: dict, status: dict, result: dict):
status['status'] = 'running'
# Get our parameters
n_neighbors = args['n_neighbors']
min_dist = args['min_dist']
metric = args['metric']
iterations = args['iterations']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.values
row_labels = df.index.values
# We need to drop any NaNs
df = df.dropna()
data = df[
columns[1:]].values # skip over the label and just pull the data
if (args['scale']):
data = StandardScaler().fit_transform(data)
reducer = umap.UMAP()
embedding = reducer.fit_transform(data)
result_df = pd.DataFrame(embedding, columns={'x', 'y'})
result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(columns)
for row in result_df.reset_index().values.tolist():
result_list.append(row)
result['embedding'] = result_list
status['status'] = 'done'
def community_detection(self, args: dict, status: dict, result: dict):
status['status'] = 'running'
# Get our parameters
n_neighbors = args['n_neighbors']
reg = args['reg']
eigen_solver = args['eigen_solver']
tol = args['tol']
max_iter = args['max_iter']
method = args['method']
hessian_tol = args['hessian_tol']
modified_tol = args['modified_tol']
neighbors_algorithm = args['neighbors_algorithm']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.to_list()
row_labels = df.index.to_list()
# We need to drop any NaNs
df = df.dropna()
data = df[
columns[1:]].values # skip over the label and just pull the data
lle = LocallyLinearEmbedding(n_components=2,
n_neighbors=n_neighbors,
reg=reg,
eigen_solver=eigen_solver,
tol=tol,
max_iter=max_iter,
method=method,
hessian_tol=hessian_tol,
modified_tol=modified_tol,
neighbors_algorithm=neighbors_algorithm,
n_jobs=10)
embedding = lle.fit_transform(data)
#print(str(embedding))
result_df = pd.DataFrame(embedding, columns=['x', 'y'])
result_df.insert(0, "Node", row_labels)
#print(str(result_df))
# result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(['Node', 'x', 'y'])
for row in result_df.reset_index().values.tolist():
result_list.append(row[1:])
result['embedding'] = result_list
status['status'] = 'done'
def community_detection(self, args: dict, status: dict, result: dict):
status['status'] = 'running'
# Get our parameters
metric = args['metric']
n_init = args['n_init']
max_iter = args['max_iter']
eps = args['eps']
dissimilarity = args['dissimilarity']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.to_list()
row_labels = df.index.to_list()
# We need to drop any NaNs
df = df.dropna()
data = df[
columns[1:]].values # skip over the label and just pull the data
mds = manifold.MDS(n_components=2,
n_init=n_init,
eps=eps,
dissimilarity=dissimilarity,
metric=metric,
max_iter=max_iter,
n_jobs=10)
embedding = mds.fit_transform(data)
#print(str(embedding))
result_df = pd.DataFrame(embedding, columns=['x', 'y'])
result_df.insert(0, "Node", row_labels)
#print(str(result_df))
# result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(['Node', 'x', 'y'])
for row in result_df.reset_index().values.tolist():
result_list.append(row[1:])
result['embedding'] = result_list
status['status'] = 'done'
def community_detection(self, args: dict, status: dict, result: dict):
status['status'] = 'running'
# Get our parameters
affinity = args['affinity']
gamma = args['gamma']
eigen_solver = args['eigen_solver']
n_neighbors = args['n_neighbors']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.to_list()
row_labels = df.index.to_list()
# We need to drop any NaNs
df = df.dropna()
data = df[
columns[1:]].values # skip over the label and just pull the data
spectral = manifold.SpectralEmbedding(n_components=2,
affinity=affinity,
gamma=gamma,
eigen_solver=eigen_solver,
n_neighbors=n_neighbors,
n_jobs=10)
embedding = spectral.fit_transform(data)
#print(str(embedding))
result_df = pd.DataFrame(embedding, columns=['x', 'y'])
result_df.insert(0, "Node", row_labels)
#print(str(result_df))
# result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(['Node', 'x', 'y'])
for row in result_df.reset_index().values.tolist():
result_list.append(row[1:])
result['embedding'] = result_list
status['status'] = 'done'
def community_detection(self, args: dict, status: dict, result: dict):
status['status'] = 'running'
# Get our parameters
solver = args['solver']
shrinkage = args['shrinkage']
tol = args['tol']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.to_list()
row_labels = df.index.to_list()
# We need to drop any NaNs
df = df.dropna()
data = df[
columns[1:]].values # skip over the label and just pull the data
lda = LinearDiscriminantAnalysis(solver=solver,
shrinkage=shrinkage,
tol=tol)
embedding = lda.fit_transform(data)
#print(str(embedding))
result_df = pd.DataFrame(embedding, columns=['x', 'y'])
result_df.insert(0, "Node", row_labels)
#print(str(result_df))
# result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(['Node', 'x', 'y'])
for row in result_df.reset_index().values.tolist():
result_list.append(row[1:])
result['embedding'] = result_list
status['status'] = 'done'
def community_detection(self, args:dict, status:dict, result:dict):
status['status'] = 'running'
# Get our parameters
perplexity = args['perplexity']
early_ex = args['early_exaggeration']
learning_rate = args['learning_rate']
n_iter = args['n_iter']
metric = args['metric']
init = args['init']
data = args['json_data']
df = utils.get_matrix(data)
columns = df.columns.to_list()
row_labels = df.index.to_list()
# We need to drop any NaNs
df = df.dropna()
data = df[columns[1:]].values # skip over the label and just pull the data
tsne = TSNE(n_components=2, perplexity=perplexity,early_exaggeration=early_ex, learning_rate=learning_rate,
metric=metric, init=init, n_jobs=10)
embedding = tsne.fit_transform(data)
#print(str(embedding))
result_df = pd.DataFrame(embedding, columns=['x','y'])
result_df.insert(0,"Node",row_labels)
#print(str(result_df))
# result_df[columns[0]] = [str(x) for x in row_labels]
result_list = []
result_list.append(['Node','x','y'])
for row in result_df.reset_index().values.tolist():
result_list.append(row[1:])
result['embedding'] = result_list
status['status'] = 'done'