aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorCalvin <calvin@EESI>2013-03-06 10:18:43 -0500
committerCalvin <calvin@EESI>2013-03-06 10:18:43 -0500
commit7f6e740ea9a3358f431c80c4db43eacbc6e8778c (patch)
treec1c1adf57e08a26e98b6f6ccbf7ea3df427ce5a6
parent991f0940eeb412b67cfe36a09701b09d03367c93 (diff)
formatting cleanups, removed unused import, deleted unused lines
-rwxr-xr-xmultifasta_to_otu.py12
1 files changed, 5 insertions, 7 deletions
diff --git a/multifasta_to_otu.py b/multifasta_to_otu.py
index 9ce5f48..f5dbe6a 100755
--- a/multifasta_to_otu.py
+++ b/multifasta_to_otu.py
@@ -5,7 +5,6 @@ from Bio import SeqIO
import multiprocessing
from subprocess import *
import os
-import quikr_train as qt
import quikr as q
import sys
import numpy as np
@@ -48,13 +47,13 @@ def main():
# Make sure our input exist
if not os.path.isdir(args.input_directory):
- parser.error( "Input directory not found")
+ parser.error("Input directory not found")
if not os.path.isdir(args.output_directory):
- parser.error( "Input directory not found")
+ parser.error("Output directory not found")
if not os.path.isdir(args.output_directory):
- os.path.mkdir(args,output_directory)
+ os.path.mkdir(args, output_directory)
if not os.path.isfile(args.trained_matrix):
parser.error("custom trained matrix not found")
@@ -70,7 +69,7 @@ def main():
kmer = args.kmer
# Load trained matrix
- trained_matrix = np.load(args.trained_matrix);
+ trained_matrix = np.load(args.trained_matrix)
# Return a list of the input directory
fasta_list = os.listdir(args.input_directory)
@@ -92,7 +91,7 @@ def main():
# load the keys with values from each fasta result
for fasta, fasta_it in map(None, fasta_list, range(len(fasta_list))):
- count_sequences = Popen(["grep", "-c" ,"^>", args.input_directory + fasta], stdout=PIPE)
+ count_sequences = Popen(["grep", "-c" , "^>", args.input_directory + fasta], stdout=PIPE)
number_of_sequences = int(count_sequences.stdout.readline())
proportions = np.loadtxt(output_directory + fasta);
@@ -121,7 +120,6 @@ def main():
#write out our fasta file row
writer.writerow(['# QIIME vGail OTU table'])
fasta_row = ['#OTU_ID']
- fasta_list
fasta_row.append(' '.join(fasta_list))
fasta_row = [' '.join(fasta_row)]
writer.writerow(fasta_row)