def _banyan(es, key_type, alg):
t = banyan.SortedSet(key_type = key_type, alg = alg)
for i, e in enumerate(es):
t.add(e)
if i % 1 == 0:
for ee in t:
pass
def build_trees(substacks):
tx = bn.SortedSet(key_type=(int, int),
updator=bn.OverlappingIntervalsUpdator)
ty = bn.SortedSet(key_type=(int, int),
updator=bn.OverlappingIntervalsUpdator)
tz = bn.SortedSet(key_type=(int, int),
updator=bn.OverlappingIntervalsUpdator)
idx = {}
for s in substacks.values():
xx = (s.info['X0'], s.info['X0'] + s.info['Width'] - 1)
tx.add(xx)
yy = (s.info['Y0'], s.info['Y0'] + s.info['Height'] - 1)
ty.add(yy)
zz = (s.info['Z0'], s.info['Z0'] + s.info['Depth'] - 1)
tz.add(zz)
idx[(xx, yy, zz)] = s.substack_id
return tx, ty, tz, idx
def _banyan(es, key_type, alg, updator):
t = banyan.SortedSet(key_type=key_type, alg=alg, updator=updator)
if updator is None:
for e in es:
t.add(e)
if len(t) > 1:
keys = list(t)
c = min([abs(x - y) for (x, y) in zip_(keys[1:], keys[:-1])])
else:
for e in es:
t.add(e)
if len(t) > 1:
c = t.min_gap()
def get_genes_in_region(self, region_start, region_end):
"""
List of gene_ids of genes that overlap this region
Inclusive
"""
if not hasattr(self, '_genetree'):
import banyan
self._genetree = banyan.SortedSet([(t[1], t[2]) for t in self.get_ordered_genes()], updator=banyan.OverlappingIntervalsUpdator)
self._geneposmap = {(t[1], t[2]): t[0] for t in self.get_ordered_genes()}
ret = []
for item in self._genetree.overlap((region_start, region_end)):
ret.append(self._geneposmap[item])
return ret
def _banyan(es, key_type, alg, updator):
t = banyan.SortedSet(key_type=key_type, alg=alg, updator=updator)
if updator is None:
for e in es:
t.add(e)
c = 0
for i in t:
if i == e:
break
c += 1
else:
for e in es:
t.add(e)
t.order(e)
def run_tests(names, num_items, num_its):
fns = dict([
('bintrees', lambda es: bintrees.FastRBTree([(e, None) for e in es])),
('set', lambda es: set(es)),
('banyan_red_black_tree',
lambda es: banyan.SortedSet(es, alg = banyan.RED_BLACK_TREE)),
('banyan_splay_tree',
lambda es: banyan.SortedSet(es, alg = banyan.SPLAY_TREE)),
('banyan_sorted_list',
lambda es: banyan.SortedSet(es, alg = banyan.SORTED_LIST)),
('banyan_red_black_tree_gen',
lambda es: banyan.SortedSet(es, key_type = int, alg = banyan.RED_BLACK_TREE)),
('banyan_splay_tree_gen',
lambda es: banyan.SortedSet(es, key_type = int, alg = banyan.SPLAY_TREE)),
('banyan_sorted_list_gen',
lambda es: banyan.SortedSet(es, key_type = int, alg = banyan.SORTED_LIST))])
t = dict([])
for name in names:
t[name] = _run_test(fns[name], int, num_items, num_its)
return t
def run_tests(names, num_items, num_its, mini_num_items = 1, type_ = int):
fns = dict([
('btrees', lambda es: BTrees.OOBTree.OOSet(es)),
('btree', lambda es: btree.sorted_btree.bulkload(sorted(es), 128)),
('blist', lambda es: blist.sortedset(es)),
('bintrees', lambda es: bintrees.FastRBTree([(e, None) for e in es])),
('set', lambda es: set(es)),
('banyan_red_black_tree',
lambda es: banyan.SortedSet(es, key_type = type_, alg = banyan.RED_BLACK_TREE)),
('banyan_splay_tree',
lambda es: banyan.SortedSet(es, key_type = type_, alg = banyan.SPLAY_TREE)),
('banyan_sorted_list',
lambda es: banyan.SortedSet(es, key_type = type_, alg = banyan.SORTED_LIST)),
('banyan_red_black_tree_gen',
lambda es: banyan.SortedSet(es, alg = banyan.RED_BLACK_TREE)),
('banyan_splay_tree_gen',
lambda es: banyan.SortedSet(es, alg = banyan.SPLAY_TREE)),
('banyan_sorted_list_gen',
lambda es: banyan.SortedSet(es, alg = banyan.SORTED_LIST))])
t = dict([])
for name in names:
t[name] = _run_test(fns[name], type_, num_items, num_its, mini_num_items)
return t
def _banyan(es, key_type, alg):
t = banyan.SortedSet(es, key_type = key_type, alg = alg)
def _banyan(es, key_type):
t = banyan.SortedSet(updator = banyan.OverlappingIntervalsUpdator, key_type = key_type)
for e in es:
t.add(e)
c = len(t.overlap(e))
def _banyan(es, key_type, alg, updator):
t = banyan.SortedSet(key_type = key_type, alg = alg, updator = updator)
for e in es:
t.add(e)
for e in es:
t.remove(e)
def _banyan(es, key_type, alg):
t = banyan.SortedSet(key_type=key_type, alg=alg)
for e in es:
t.add(e)
if node is None:
return
_fancy_trace(node.right, path + 'r')
node_content = ''
if path:
for d, nd in itertools.izip(path[:-1], path[1:]):
node_content += ' ' if d == nd else '| '
node_content += '/-' if path[-1] == 'r' else '\-'
node_content += str(node.key)
if node.metadata is not None:
node_content += ' [ ' + str(node.metadata) + ' ]'
print(node_content)
_fancy_trace(node.left, path + 'l')
_fancy_trace(node, '')
if __name__ == '__main__':
keys = [random.randint(1, 100) for i in range(18)]
simple_trace(banyan.SortedSet(keys).root)
fancy_trace(banyan.SortedSet(keys).root)
fancy_trace(banyan.SortedSet(keys, updator=banyan.RankUpdator).root)
fancy_trace(banyan.SortedSet(keys, updator=banyan.MinMaxUpdator).root)