def buildIndex(K, inputs, output):
"""
Create a new k-mer index. The FASTA files named in the list
`inputs` are read in and the `K` length k-mers and their reverse
complements are extracted and collated to create an index that
maps from k-mer to sequence number (numbering from 0). The
`names` member of the KmerIndex object can be used to retrieve
the name from the sequence number.
"""
seqs = []
for inp in inputs:
with openFile(inp) as f:
seqs += list(readFasta(f))
S = []
nms = []
lens = array.array('I', [])
for i in xrange(len(seqs)):
(nm, seq) = seqs[i]
nms.append(nm)
xs = list(kmers(K, seq, True))
xs.sort()
uniq(xs)
seqs[i] = [nm, xs]
lens.append(len(xs))
S += xs
S.sort()
uniq(S)
S = sparse(2*K, S)
T = array.array('I', [0 for i in xrange(S.count() + 1)])
for i in xrange(len(seqs)):
for x in seqs[i][1]:
r = S.rank(x)
T[r] += 1
t0 = 0
for i in xrange(len(T)):
t1 = t0 + T[i]
T[i] = t0
t0 = t1
T0 = [c for c in T]
U = array.array('H', [0 for i in xrange(t0)])
for i in xrange(len(seqs)):
for x in seqs[i][1]:
r = S.rank(x)
U[T0[r]] = i
T0[r] += 1
with container(output, 'w') as z:
writeKmers(K, S.xs, z)
n = write32(z, T, 'offsets')
z.meta['T'] = n
n = write16(z, U, 'postings')
z.meta['U'] = n
n = write32(z, lens, 'lens')
z.meta['lens'] = n
z.meta['names'] = nms
def test_uniq_few():
xs = [1, 1, 2, 3, 3, 3, 4]
misc.uniq(xs)
assert len(xs) == 4
assert xs[0] == 1
assert xs[1] == 2
assert xs[2] == 3
assert xs[3] == 4
def test_uniq_simple():
xs = [i for i in xrange(10)]
print xs
misc.uniq(xs)
print xs
assert len(xs) == 10
for i in xrange(10):
assert xs[i] == i
def test_uniq_many():
xs = []
ys = set([])
random.seed(17)
N = 100000
for i in xrange(N):
x = random.randint(0, 1000)
xs.append(x)
ys.add(x)
ys = list(ys)
ys.sort()
xs.sort()
misc.uniq(xs)
assert len(xs) == len(ys)
for i in xrange(len(xs)):
assert xs[i] == ys[i]
def neigh(K, x, d):
if d == 0:
return []
xs = []
for i in xrange(K):
for j in xrange(3):
xs.append(x ^ ((j + 1) << (2 * i)))
xs.sort()
if d == 1:
return xs
zs = []
for y in xs:
zs += neigh(K, y, d - 1)
zs.sort()
uniq(zs)
diff(zs, [x])
diff(zs, xs)
return zs
from pykmer.basics import kmers
from pykmer.file import readFasta
from pykmer.misc import uniq
import pykmer.kset as kset
import sys
if len(sys.argv) < 4:
print >> sys.stderr, "usage: make-kset.py <K> <output-filename> <input-FASTA>...."
sys.exit(1)
K = int(sys.argv[1])
xs = []
for fn in sys.argv[3:]:
with open(fn) as f:
for (nm, seq) in readFasta(f):
xs += list(kmers(K, seq, True))
xs.sort()
uniq(xs)
kset.write(K, xs, sys.argv[2])
def test_uniq_empty():
xs = []
misc.uniq(xs)
assert len(xs) == 0
def test_uniq_singleton():
xs = [1]
misc.uniq(xs)
assert len(xs) == 1
assert xs[0] == 1
