return args
# MAIN
if __name__ == "__main__":
args = interface()
input_dir = os.path.abspath(args.IN)
if not input_dir.endswith('/'):
input_dir += '/'
output_dir = os.path.abspath(args.OUT)
if not output_dir.endswith('/'):
output_dir += '/'
thresh = args.threshold
hashobject = StreamingEigenhashes(input_dir, output_dir, get_pool=-1)
Kmer_Hash_Count_Files = glob.glob(os.path.join(hashobject.input_path, '*.count.hash.conditioned'))
hashobject.path_dict = {}
for i in range(len(Kmer_Hash_Count_Files)):
hashobject.path_dict[i] = Kmer_Hash_Count_Files[i]
lsi = models.LsiModel.load(hashobject.output_path + 'kmer_lsi.gensim')
hashobject.cluster_thresh = thresh
Index = hashobject.lsi_cluster_index(lsi)
np.save(hashobject.output_path + 'cluster_index.npy', Index)
print('Cluster index has shape: ' + str(Index.shape))
with open(hashobject.output_path + 'numClusters.txt', 'w') as f:
f.write('{0}\n'.format(Index.shape[0]))
import numpy as np
from streaming_eigenhashes import StreamingEigenhashes
help_message = 'usage example: python kmer_clusters.py -i /project/home/hashed_reads/ -o /project/home/cluster_vectors/'
if __name__ == "__main__":
try:
opts, args = getopt.getopt(sys.argv[1:], 'hi:o:',
["inputdir=", "outputdir="])
except:
print help_message
sys.exit(2)
for opt, arg in opts:
if opt in ('-h', '--help'):
print help_message
sys.exit()
elif opt in ('-i', '--inputdir'):
inputdir = arg
if inputdir[-1] != '/':
inputdir += '/'
elif opt in ('-o', '--outputdir'):
outputdir = arg
if outputdir[-1] != '/':
outputdir += '/'
hashobject = StreamingEigenhashes(inputdir, outputdir, get_pool=False)
Kmer_Hash_Count_Files = glob.glob(
os.path.join(hashobject.input_path, '*.nonzero.npy'))
corpus_generator = hashobject.corpus_idf_from_hash_paths(
Kmer_Hash_Count_Files)
hashobject.train_tfidf(corpus_generator)
np.save(hashobject.output_path + 'global_weights.npy',
hashobject.global_weights)
if opt in ('-h', '--help'):
print help_message
sys.exit()
elif opt in ('-i', '--inputdir'):
inputdir = arg
if inputdir[-1] != '/':
inputdir += '/'
elif opt in ('-o', '--outputdir'):
outputdir = arg
if outputdir[-1] != '/':
outputdir += '/'
elif opt in ('-p', '--numproc'):
num_proc = int(arg)
elif opt in ('-s', '--single'):
singleInstance = True
### use -p option for multiprocessing
num_proc = -1
###
hashobject = StreamingEigenhashes(inputdir, outputdir, get_pool=num_proc)
Kmer_Hash_Count_Files = glob.glob(
os.path.join(hashobject.input_path, '*.count.hash.conditioned'))
hashobject.path_dict = {}
for i in range(len(Kmer_Hash_Count_Files)):
hashobject.path_dict[i] = Kmer_Hash_Count_Files[i]
corpus = hashobject.kmer_corpus_from_disk()
# This is a hack. Should do a better job chosing num_dims
lsi = hashobject.train_kmer_lsi(corpus,
num_dims=len(hashobject.path_dict) * 4 / 5,
single=singleInstance)
lsi.save(hashobject.output_path + 'kmer_lsi.gensim')
