def main():
input_fname = sys.argv[1]
out_fname = sys.argv[2]
reader = BedGraphReader( open( input_fname ) )
# Fill array from reader
d = array_tree_dict_from_reader( reader, {}, block_size=BLOCK_SIZE )
for array_tree in d.itervalues():
array_tree.root.build_summary()
FileArrayTreeDict.dict_to_file( d, open( out_fname, "w" ) )
def setUp(self):
tree = ArrayTree(10000, 10) # max value of 10000, each block has 10 numbers
for i in range(5000):
tree[i] = i
# Insert extra copies to test frequency
for i in range(3000):
tree[i] = i
tree.set_range(5000, 9001, 100)
tree.root.build_summary()
d = {'test': tree}
f = tempfile.TemporaryFile()
FileArrayTreeDict.dict_to_file(d, f)
f.seek(0)
self.filearraytreedict = FileArrayTreeDict(f)
self.filearraytree = self.filearraytreedict['test']
def setUp(self):
tree = ArrayTree(10000, 10) # max value of 10000, each block has 10 numbers
for i in range(5000):
tree[i] = i
# Insert extra copies to test frequency
for i in range(3000):
tree[i] = i
tree.set_range(5000, 9001, 100)
tree.root.build_summary()
d = {'test': tree}
f = tempfile.TemporaryFile()
FileArrayTreeDict.dict_to_file( d, f )
f.seek(0)
self.filearraytreedict = FileArrayTreeDict(f)
self.filearraytree = self.filearraytreedict['test']
def main():
sizes_fname = sys.argv[1]
out_fname = sys.argv[2]
sizes = {}
for line in open(sizes_fname):
fields = line.split()
sizes[fields[0]] = int(fields[1])
# Fill array from wiggle
d = array_tree_dict_from_reader(WiggleReader(sys.stdin), sizes)
for value in d.values():
value.root.build_summary()
with open(out_fname, "w") as f:
FileArrayTreeDict.dict_to_file(d, f)
def test_big(self):
tree = ArrayTree(2147483647, 1000) # What we use for tracks
for i in range(5000):
tree[i] = i
# Insert extra copies to test frequency
for i in range(3000):
tree[i] = i
tree.set_range(5000, 9001, 100)
tree.set_range(14000000, 15000000, 200)
tree.root.build_summary()
d = {'test': tree}
f = tempfile.TemporaryFile()
FileArrayTreeDict.dict_to_file(d, f)
f.seek(0)
at = FileArrayTreeDict(f)['test']
lvl1 = at.get_summary(14000000, 1)
avgs = [float(_) for _ in lvl1.sums/lvl1.counts]
self.assertEqual(len(avgs), 1000)
self.assertEqual(avgs, [200 for i in range(0, 1000)])
def test_big(self):
tree = ArrayTree(2147483647, 1000) # What we use for tracks
for i in range(5000):
tree[i] = i
# Insert extra copies to test frequency
for i in range(3000):
tree[i] = i
tree.set_range(5000, 9001, 100)
tree.set_range(14000000, 15000000, 200)
tree.root.build_summary()
d = {'test': tree}
f = tempfile.TemporaryFile()
FileArrayTreeDict.dict_to_file( d, f )
f.seek(0)
at = FileArrayTreeDict(f)['test']
lvl1 = at.get_summary(14000000, 1)
avgs = [float(_) for _ in lvl1.sums/lvl1.counts]
self.assertEqual( len(avgs), 1000 )
self.assertEqual( avgs, [ 200 for i in range(0, 1000)] )
def get_stats(self, chrom):
f = open(self.dataset.file_name)
d = FileArrayTreeDict(f)
try:
chrom_array_tree = d[chrom]
except KeyError:
return "no data"
root_summary = chrom_array_tree.get_summary(0, chrom_array_tree.levels)
f.close()
return {
'max': float(max(root_summary.maxs)),
'min': float(min(root_summary.mins))
}
def get_data(self, chrom, start, end, **kwargs):
f = open(self.dataset.file_name)
d = FileArrayTreeDict(f)
# Get the right chromosome
try:
chrom_array_tree = d[chrom]
except KeyError:
return None
block_size = chrom_array_tree.block_size
start = int(start)
end = int(end)
resolution = max(1, ceil(float(kwargs['resolution'])))
level = int(floor(log(resolution, block_size)))
level = max(level, 0)
stepsize = block_size**level
step1 = stepsize * block_size
# Is the requested level valid?
assert 0 <= level <= chrom_array_tree.levels
results = []
for block_start in range(start, end, stepsize * block_size):
# print block_start
# Return either data point or a summary depending on the level
indexes = range(block_start, block_start + stepsize * block_size,
stepsize)
if level > 0:
s = chrom_array_tree.get_summary(block_start, level)
if s is not None:
results.extend(zip(indexes, map(float, s.sums / s.counts)))
else:
v = chrom_array_tree.get_leaf(block_start)
if v is not None:
results.extend(zip(indexes, map(float, v)))
f.close()
return results
