def demo_LdaCgsSeq(doc_len=500,
V=100000,
n_docs=100,
K=20,
n_iterations=5,
corpus_seed=None,
model_seed=None):
from vsm.extensions.corpusbuilders import random_corpus
print('Words per document:', doc_len)
print('Words in vocabulary:', V)
print('Documents in corpus:', n_docs)
print('Number of topics:', K)
print('Iterations:', n_iterations)
c = random_corpus(n_docs * doc_len,
V,
doc_len,
doc_len + 1,
seed=corpus_seed)
m = LdaCgsSeq(c, 'document', K=K, seed=model_seed)
m.train(n_iterations=n_iterations, verbose=2)
return m
def setUp(self):
from vsm.extensions.corpusbuilders import random_corpus
c = random_corpus(1000, 100, 0, 20)
self.m = BeagleEnvironment(c, n_cols=100)
self.m.train()
def setUp(self):
from vsm.extensions.corpusbuilders import random_corpus
c = random_corpus(1000, 100, 0, 20)
self.m = BeagleEnvironment(c, n_cols=100)
self.m.train()
def demo_LdaCgsSeq(doc_len=500, V=100000, n_docs=100,
K=20, n_iterations=5,
corpus_seed=None, model_seed=None):
from vsm.extensions.corpusbuilders import random_corpus
print 'Words per document:', doc_len
print 'Words in vocabulary:', V
print 'Documents in corpus:', n_docs
print 'Number of topics:', K
print 'Iterations:', n_iterations
c = random_corpus(n_docs*doc_len, V, doc_len, doc_len+1, seed=corpus_seed)
m = LdaCgsSeq(c, 'document', K=K, seed=model_seed)
m.train(n_iterations=n_iterations, verbose=2)
return m
def demo_LdaCgsMulti(doc_len=500, V=100000, n_docs=100,
K=20, n_iterations=5, n_proc=2,
corpus_seed=None, model_seeds=None):
from vsm.extensions.corpusbuilders import random_corpus
print('Words per document:', doc_len)
print('Words in vocabulary:', V)
print('Documents in corpus:', n_docs)
print('Number of topics:', K)
print('Iterations:', n_iterations)
print('Number of processors:', n_proc)
c = random_corpus(n_docs*doc_len, V, doc_len, doc_len+1, seed=corpus_seed)
m = LdaCgsMulti(c, 'document', K=K, n_proc=n_proc, seeds=model_seeds)
m.train(n_iterations=n_iterations, verbose=2)
return m
def demo_LdaCgs(doc_len=500, V=100000, n_docs=100,
K=20, n_iterations=5, n_threads=1):
from vsm.extensions.corpusbuilders import random_corpus
print('Words per document:', doc_len)
print('Words in vocabulary:', V)
print('Documents in corpus:', n_docs)
print('Number of topics:', K)
print('Iterations:', n_iterations)
c = random_corpus(n_docs*doc_len, V, doc_len, doc_len+1)
print('Random corpus generated. Initializing model.')
m = LdaCgs(c, 'document', K=K)
print('Begin estimation.')
m.train(n_iterations=n_iterations, n_threads=n_threads)
return m
def demo_LdaCgs(doc_len=500, V=100000, n_docs=100,
K=20, n_iterations=5, n_threads=1):
from vsm.extensions.corpusbuilders import random_corpus
print 'Words per document:', doc_len
print 'Words in vocabulary:', V
print 'Documents in corpus:', n_docs
print 'Number of topics:', K
print 'Iterations:', n_iterations
c = random_corpus(n_docs*doc_len, V, doc_len, doc_len+1)
print 'Random corpus generated. Initializing model.'
m = LdaCgs(c, 'document', K=K)
print 'Begin estimation.'
m.train(n_iterations=n_iterations, n_threads=n_threads)
return m
