def sample_matrix_to_runs(dist, reps=3):
'''Repeats a distance matrix to expand samples to reps.'''
runs = DistanceMatrix(
np.repeat(np.repeat(dist.data, reps, axis=1), reps, axis=0))
runs.ids = [
'{}-{}'.format(g, i + 1) for g in dist.ids for i in range(reps)
]
return runs
def get_truth(treefile, reps):
dist = partition_weighted_distance(treefile)
runs = DistanceMatrix(
np.repeat(np.repeat(dist.data, reps, axis=1), reps, axis=0))
runs.ids = [
'{}-{}'.format(g, i + 1) for g in dist.ids for i in range(reps)
]
return runs
def go_distance(args):
#print args
try:
json_data = open(args.in_file)
data = json.load(json_data)
json_data.close()
except:
print("NO FILE FOUND ERROR")
sys.exit()
datasets = []
for i in data['columns']:
#print i['id']
datasets.append(i['id'])
z = np.array(data['data'])
#dmatrix = np.transpose(z)
(dmatrix, bad_rows) = remove_zero_sum_datasets(np.transpose(z))
# find zero sum rows (datasets) after transpose
#print(dmatrix)
# delete datasets too:
edited_dataset_list=[]
#edited_did_hash = {}
for row,line in enumerate(data['columns']):
if row not in bad_rows[0]:
edited_dataset_list.append(line['id'])
#print(edited_dataset_list)
dist = get_dist(dmatrix, data)
if args.create_splits:
# we're done
return
#print(dist)
#sys.exit()
dm1 = get_dist_matrix1(dist)
dm2 = {}
dm3 = {}
for row,name in enumerate(edited_dataset_list):
name = str(name)
dm2[name] = {}
#file_data_line = name+','
for col,d in enumerate(dm1[row]):
#print data['columns'][col]['id']
#file_data_line += str(dm1[row][col])+','
dm2[name][str(data['columns'][col]['id'])] = dm1[row][col]
dm3[(name, str(data['columns'][col]['id']))] = dm1[row][col]
#file_data_line = file_data_line[:-1]+'\n'
#out_fp.write(file_data_line)
out_file_selected = os.path.join(args.basedir, 'tmp', args.prefix+'_distance.json')
#my_file = Path(out_file_selected)
if not os.path.exists(out_file_selected):
out_fp2 = open(out_file_selected,'w')
out_fp2.write(json.dumps(dm2))
out_fp2.close()
dm1 = DistanceMatrix(dm1) # convert to scikit-bio DistanceMatrix (v 0.5.1)
dm1.ids = edited_dataset_list # assign row names
#print(dm1)
return (dm1, edited_dataset_list)
def test_embad(self):
exp = DistanceMatrix([[0, 2, 4, 6], [2, 0, 2, 4], [4, 2, 0, 2],
[6, 4, 2, 0]])
exp.ids = ['S1', 'S2', 'S3', 'S4']
res = embad(self.table)
self.assertEquals(exp, res)
def get_truth(treefile, reps):
dist = partition_weighted_distance(treefile)
runs = DistanceMatrix(
np.repeat(np.repeat(dist.data, reps, axis=1), reps, axis=0))
runs.ids = ['{}-{}'.format(g, i+1) for g in dist.ids for i in range(reps)]
return runs
def sample_matrix_to_runs(dist, reps=3):
'''Repeats a distance matrix to expand samples to reps.'''
runs = DistanceMatrix(
np.repeat(np.repeat(dist.data, reps, axis=1), reps, axis=0))
runs.ids = ['{}-{}'.format(g, i+1) for g in dist.ids for i in range(reps)]
return runs
def go_distance(args):
#print args
try:
json_data = open(args.in_file)
data = json.load(json_data)
json_data.close()
except:
print("1-NO FILE FOUND ERROR")
sys.exit()
datasets = []
for i in data['columns']:
#print i['id']
datasets.append(i['id'])
z = np.array(data['data'])
#dmatrix = np.transpose(z)
(dmatrix, bad_rows) = remove_zero_sum_datasets(np.transpose(z))
# find zero sum rows (datasets) after transpose
#print(dmatrix)
# delete datasets too:
edited_dataset_list = []
#edited_did_hash = {}
for row, line in enumerate(data['columns']):
if row not in bad_rows[0]:
edited_dataset_list.append(line['id'])
#print(edited_dataset_list)
dist = get_dist(dmatrix, data)
if args.create_splits:
# we're done
return
#print(dist)
#sys.exit()
dm1 = get_dist_matrix1(dist)
dm2 = {}
dm3 = {}
for row, name in enumerate(edited_dataset_list):
name = str(name)
dm2[name] = {}
#file_data_line = name+','
for col, d in enumerate(dm1[row]):
#print data['columns'][col]['id']
#file_data_line += str(dm1[row][col])+','
dm2[name][str(data['columns'][col]['id'])] = dm1[row][col]
dm3[(name, str(data['columns'][col]['id']))] = dm1[row][col]
#file_data_line = file_data_line[:-1]+'\n'
#out_fp.write(file_data_line)
out_file_selected = os.path.join(args.basedir,
args.prefix + '_distance.json')
out_file_csv = os.path.join(args.basedir, args.prefix + '_distance.csv')
print(dm1)
# must over write each time
#if not os.path.exists(out_file_selected):
out_fp2 = open(out_file_selected, 'w')
out_fp2.write(json.dumps(dm2))
out_fp2.close()
try:
os.chmod(out_file_selected, 0o664)
except:
pass
write_csv_file(dm1, data, args)
dm1 = DistanceMatrix(dm1) # convert to scikit-bio DistanceMatrix (v 0.5.1)
dm1.ids = edited_dataset_list # assign row names
#print(dm1)
return (dm1, edited_dataset_list)
