def fqframe(fileh):
final_schema = Schema({
'id': str,
'seq': str,
'quality': str,
'qual_ints': check_np_type('int64'),
'error': check_np_type('float64'),
'description': str
})
#get_object = _id
index = ['id']
columns = ('id', 'seq', 'quality', 'description', 'qual_ints', 'error')
SANGER = True
get_id = attr('id')
get_seq = compose(str, attr('seq'))
get_qual_ints = compose_all(np.array, itemgetter('phred_quality'),
attr('_per_letter_annotations'))
get_description = attr('description')
get_quality = SeqIO.QualityIO._get_sanger_quality_str
get_error = compose(error, get_qual_ints)
#get_error = error_from_ints(get_qual_ints)
getters = [
get_id, get_seq, get_quality, get_description, get_qual_ints, get_error
]
assert len(getters) == len(columns)
metadata = {'filename': fileh.name}
iterator = get_fastq(fileh)
get_raw_record = partial(next, iterator)
# def get_row(record):
# #record = next(fileh)
## import sys
## __module__ = sys.modules[__name__]
## get_getter = compose(attr, "get_{0}".format)
## _getters = map(get_getter, columns)
## self_getters = apply_each(_getters, __module__) #fzip(_getters, repeat(__module__, clen))
# results = apply_each(self_getters, record)
# final_dict = dict(zip(columns, results))
# final_schema.validate(final_dict)
# return final_dict
# def load_fastq():
# fq = get_fastq(fileh)
# dicts = map(get_row, fq)
# return pd.DataFrame(dicts).set_index(index) #, index=index, columns=columns)
#jreturn nameddict(
return {
'obj_func': get_raw_record,
'columns': columns,
'getters': getters,
'validator': final_schema,
'dictgetters': None
}
class Seq(object)
#SANGER_OFFSET = 33
SANGER = True
get_seq_str = compose(str, attr('seq'))
get_id = attr('id')
get_qual_ints = compose(itemgetter('phred_quality'), attr('_per_letter_annotations'))
get_description = attr('description')
phred_to_char = chr if not SANGER else compose(chr, lambda a: a - 33)
get_qual_chars = compose(pmap(phred_to_char), get_qual_ints)
index = 'id'
_object = _id
def fqframe(fileh):
final_schema = Schema({
'id' : str,
'seq' : str,
'quality' : str,
'qual_ints' : check_np_type('int64'),
'error' : check_np_type('float64'),
'description' : str
})
#get_object = _id
index = ['id']
columns = ('id', 'seq', 'quality', 'description', 'qual_ints', 'error')
SANGER = True
get_id = attr('id')
get_seq= compose(str, attr('seq'))
get_qual_ints = compose_all(np.array, itemgetter('phred_quality'), attr('_per_letter_annotations'))
get_description = attr('description')
get_quality = SeqIO.QualityIO._get_sanger_quality_str
get_error = compose(error, get_qual_ints)
#get_error = error_from_ints(get_qual_ints)
getters = [get_id, get_seq, get_quality, get_description, get_qual_ints, get_error]
assert len(getters) == len(columns)
metadata = {'filename' : fileh.name}
iterator = get_fastq(fileh)
get_raw_record = partial(next, iterator)
# def get_row(record):
# #record = next(fileh)
## import sys
## __module__ = sys.modules[__name__]
## get_getter = compose(attr, "get_{0}".format)
## _getters = map(get_getter, columns)
## self_getters = apply_each(_getters, __module__) #fzip(_getters, repeat(__module__, clen))
# results = apply_each(self_getters, record)
# final_dict = dict(zip(columns, results))
# final_schema.validate(final_dict)
# return final_dict
# def load_fastq():
# fq = get_fastq(fileh)
# dicts = map(get_row, fq)
# return pd.DataFrame(dicts).set_index(index) #, index=index, columns=columns)
#jreturn nameddict(
return { 'obj_func' : get_raw_record,
'columns' : columns,
'getters' : getters,
'validator' : final_schema,
'dictgetters' : None
}
def run():
while len(teams['immune_system']) > 0 and len(teams['infection']) > 0:
fights = {}
unit_count = total_units()
for team in teams:
enemies = teams['immune_system'] if team == 'infection' else teams[
'infection']
for attacker in sorted(teams[team], reverse=True):
if attacker.units == 0:
continue
targets = sorted(
[(attacker.attack(enemy, simulate=True), enemy)
for enemy in enemies
if enemy not in fights.values() and enemy.units > 0],
reverse=True)
if len(targets) > 0 and targets[0][0] > 0:
fights[attacker] = targets[0][1]
for attacker in sorted(fights.keys(),
key=attr('initiative'),
reverse=True):
if attacker.units > 0:
attacker.attack(fights[attacker])
for team in teams:
to_del = []
for i, squad in enumerate(teams[team]):
if squad.units == 0:
to_del += [i]
for i in sorted(to_del, reverse=True):
del teams[team][i]
if total_units() == unit_count:
break
return len(teams['infection']) == 0
def link_list_view(request):
links = Link.objects.all().order_by('title')
files = File.objects.all().order_by('title')
link_tags = Tag.objects.all().filter(link__isnull=False).distinct()
file_tags = Tag.objects.all().filter(file__isnull=False).distinct()
tags_unsorted = list(chain(link_tags, file_tags))
tags = sorted(set(tags_unsorted), key=attr('title'))
query = request.GET.get("q")
if query:
files = pt_files.filter(
Q(title__icontains=query) | Q(text__icontains=query)
| Q(language__icontains=query)).distinct()
links = pt_links.filter(
Q(title__icontains=query) | Q(text__icontains=query)
| Q(language__icontains=query)).distinct()
tags = pt_tags.filter(
Q(title__icontains=query) | Q(text__icontains=query)
| Q(language__icontains=query)).distinct()
return render(request, 'resources/resources_list.html', {
'links': links,
'tags': tags,
'files': files
})
def find_optimal_position(bots):
box_size = 1
while box_size < max(bots, key=attr('x')).x - min(bots, key=attr('x')).x:
box_size *= 2
box_center = ORIGIN
while box_size > 0:
best, best_count = None, None
for p in points_to_test(box_center, box_size):
count = p.reaches(bots, box_size)
if best is None \
or count > best_count \
or count == best_count and p.displacement < best.displacement:
best, best_count = p, count
box_center = best
box_size //= 2
return box_center.displacement
def fqframe(fileh):
final_schema = Schema({
'id': str,
'seq': str,
'quality': str,
'qual_ints': check_np_type('int64'),
'error': check_np_type('float64'),
'description': str
})
#get_object = _id
index = ['id']
columns = ('id', 'seq', 'quality', 'description', 'qual_ints', 'error')
SANGER = True
get_id = attr('id')
get_seq = compose(str, attr('seq'))
get_qual_ints = compose_all(np.array, itemgetter('phred_quality'),
attr('_per_letter_annotations'))
get_description = attr('description')
get_quality = SeqIO.QualityIO._get_sanger_quality_str
get_error = compose(error, get_qual_ints)
#get_error = error_from_ints(get_qual_ints)
def get_row(record):
#record = next(fileh)
print(get_funcs())
import sys
__module__ = sys.modules[__name__]
get_getter = compose(attr, "get_{0}".format)
_getters = map(get_getter, columns)
self_getters = apply_each(
_getters, __module__) #fzip(_getters, repeat(__module__, clen))
results = apply_each(self_getters, record)
final_dict = dict(zip(columns, results))
final_schema.validate(final_dict)
return final_dict
def load_fastq():
fq = get_fastq(fileh)
dicts = map(get_row, fq)
return pd.DataFrame(dicts).set_index(
index) #, index=index, columns=columns)
return namedtuple('FastqFrame', ['get_row', 'load_fastq'])(
get_row, load_fastq) #{'get_row' : get_row, 'load_fastq' : load_fastq}
def started_at(path_sets, explored):
more = lazy(
filter(lambda p: p.lead_to not in explored,
(path.extend(m) for path in path_sets for m in moves)))
yield from more
yield from started_at(more, explored.union(set(map(attr("lead_to"),
more))))
def started_at(path_sets, explored):
more = lazy(filter(
lambda p: p.lead_to not in explored,
(path.extend(m) for path in path_sets for m in moves)
))
yield from more
yield from started_at(more, explored.union(set(map(attr("lead_to"), more))))
def assertSeqRecordEqual(self, seq1, seq2):
'''This is necessary because the __eq__ in SeqRecord is weird.'''
_fields = ['id', 'name', 'description']
seqstr = compose(str, attr('seq'))
self.assertEquals(seq1.letter_annotations, seq2.letter_annotations)
self.assertEquals(seqstr(seq1), seqstr(seq2))
for field in _fields:
f1, f2 = getattr(seq1, field), getattr(seq2, field)
#TODO: include field name in message?
self.assertEquals(f1, f2) #, msg=)
def fqframe(fileh):
final_schema = Schema({
'id' : str,
'seq' : str,
'quality' : str,
'qual_ints' : check_np_type('int64'),
'error' : check_np_type('float64'),
'description' : str
})
#get_object = _id
index = ['id']
columns = ('id', 'seq', 'quality', 'description', 'qual_ints', 'error')
SANGER = True
get_id = attr('id')
get_seq= compose(str, attr('seq'))
get_qual_ints = compose_all(np.array, itemgetter('phred_quality'), attr('_per_letter_annotations'))
get_description = attr('description')
get_quality = SeqIO.QualityIO._get_sanger_quality_str
get_error = compose(error, get_qual_ints)
#get_error = error_from_ints(get_qual_ints)
def get_row(record):
#record = next(fileh)
print(get_funcs())
import sys
__module__ = sys.modules[__name__]
get_getter = compose(attr, "get_{0}".format)
_getters = map(get_getter, columns)
self_getters = apply_each(_getters, __module__) #fzip(_getters, repeat(__module__, clen))
results = apply_each(self_getters, record)
final_dict = dict(zip(columns, results))
final_schema.validate(final_dict)
return final_dict
def load_fastq():
fq = get_fastq(fileh)
dicts = map(get_row, fq)
return pd.DataFrame(dicts).set_index(index) #, index=index, columns=columns)
return namedtuple('FastqFrame', ['get_row', 'load_fastq'])(get_row, load_fastq)#{'get_row' : get_row, 'load_fastq' : load_fastq}
def calculate_sky_message(particles):
height = abs(max(particles).y - min(particles).y)
time = 0
while height > 10:
for particle in particles:
particle.travel()
height = abs(max(particles).y - min(particles).y)
time += 1
ymax, ymin = max(particles).y, min(particles).y
xmax = max(particles, key=attr('x')).x
xmin = min(particles, key=attr('x')).x
try:
return time, tesseract_parse(particles, ymax, ymin, xmax, xmin)
except ImportError:
for y in range(ymin, ymax + 1):
l = ''
for x in range(xmin, xmax + 1):
l += '█' if (x, y) in particles else ' '
print(l)
print()
return time, None
def main():
scheme = Schema(
{ '<fasta>' : os.path.isfile,
Optional('--gb-file') : Or(os.path.isfile, lambda x: x is None),
Optional('--tab-file') : Or(os.path.isfile, lambda x: x is None),
Optional('--gb-id') : Or(str, lambda x: x is None),
})
raw_args = docopt(__doc__, version='Version 1.0')
args = scheme.validate(raw_args)
fasta = parse_fasta(args['<fasta>'])
genes, cds = get_genes(args['--gb-id'], args['--gb-file'], args['--tab-file'])
infos = map(partial(get_gene_degen_overlap_info, genes), map(attr('seq'), fasta))
#need `list` to force evaluation of `print`
list(map(print, map(pretty_table, infos)))
def adjust_mini(icon_label, value_label, place: int, pos: Vector):
smaller = min(icon_label, value_label, key = attr('content_width'))
if smaller is icon_label:
larger = value_label
else:
larger = icon_label
larger.x = pos.x + place * CardDrawer.PADDING.x
larger.anchor_x = 'left' if place == 1 else 'right'
smaller.x = larger.x + place * larger.content_width // 2
smaller.anchor_x = 'center'
icon_label.y = pos.y + place * CardDrawer.PADDING.y
value_label.y = pos.y - (icon_label.document.get_font().ascent
* place * (0.9 if place == 1 else 0.8))
def test_flags(self):
df = self.result
middle_e = pd.Series([False, False, False, True], dtype=object).values
outer_e = pd.Series([False, False, True, True], dtype=object).values
names = [
"template having multiple segments in sequencing",
"each segment properly aligned according to the aligner",
"segment unmapped",
"next segment in the template unmapped",
]
# inner_actual = df[df['QNAME'] == 'read2'][names]
# outer_actual1= df[df['QNAME'] == 'read1'][names]
# outer_actual3= df[df['QNAME'] == 'read3'][names]
flag8, flag60_1, flag60_2 = map(attr("values"), map(itemgetter(names), itemgetter(1, 0, 2)(df.ix)))
assert_array_equal(middle_e, flag8)
assert_array_equal(outer_e, flag60_1)
assert_array_equal(outer_e, flag60_2)
def test_flags(self):
df = self.result
middle_e = pd.Series([False, False, False, True], dtype=object).values
outer_e = pd.Series([False, False, True, True], dtype=object).values
names = [
"template having multiple segments in sequencing",
"each segment properly aligned according to the aligner",
"segment unmapped", "next segment in the template unmapped"
]
# inner_actual = df[df['QNAME'] == 'read2'][names]
# outer_actual1= df[df['QNAME'] == 'read1'][names]
# outer_actual3= df[df['QNAME'] == 'read3'][names]
flag8, flag60_1, flag60_2 = map(
attr('values'), map(itemgetter(names),
itemgetter(1, 0, 2)(df.ix)))
assert_array_equal(middle_e, flag8)
assert_array_equal(outer_e, flag60_1)
assert_array_equal(outer_e, flag60_2)
def _scale(entries):
images = [e.best_post for e in entries]
posts = []
if images:
for im in images:
im.scale = g.column_width
sm = sorted(images, key=attr('score'), reverse=True)
for im in sm[:max(1, int(len(sm) / current_app.config['AIP_PER']))]:
im.scale = g.gutter + 2 * g.column_width
for im in images:
preview_height = int(im.scale * im.height / im.width)
preview_width = int(im.scale)
posts.append(Post(
url=im.post_url,
preview_url=im.preview_url,
preview_height=preview_height,
preview_width=preview_width,
md5=im.md5
))
return posts
def sigma_scaling_selection(self, population):
fitnesses = map(attr('fitness'), population)
mu = np.mean(fitnesses)
sigma = np.std(fitnesses)
mu_sum = np.sum((fitnesses - mu)/(2*sigma) + 1)
new_population = []
for i in range(len(population)):
current_counter = 0
candidate_random_number = random.random() * mu_sum
for _ in range(1):
parents = []
for candidate in population:
current_counter += (1 + (candidate.fitness - mu) / (2 * sigma))
if current_counter >= candidate_random_number:
parents.append(candidate)
break
new_population.append(parents[0].crossover(parents[1]))
return population
def sigma_scaling_selection(self, population):
fitnesses = map(attr('fitness'), population)
mu = np.mean(fitnesses)
sigma = np.std(fitnesses)
mu_sum = np.sum((fitnesses - mu) / (2 * sigma) + 1)
new_population = []
for i in range(len(population)):
current_counter = 0
candidate_random_number = random.random() * mu_sum
for _ in range(1):
parents = []
for candidate in population:
current_counter += (1 + (candidate.fitness - mu) /
(2 * sigma))
if current_counter >= candidate_random_number:
parents.append(candidate)
break
new_population.append(parents[0].crossover(parents[1]))
return population
def leaderboard():
scores = models.UserProgress.objects.filter(score__gt=0).order_by("-score")
transforms = {_("Name"): attr("user.name"), **FN_BASE}
return {
"leaderboard": prepare_dataframe(scores, transforms, class_="table")
}
''' dict to named tuple '''
names, values = unzip(_dict.items())
return namedtuple(Name, names)(*values)
ppartial = partial(partial)
apply_to_object = compose(apply, ppartial)
kstarcompose2 = lambda f, g: lambda x: f(**g(x))
def kstarcompose(*funcs):
return reduce(kstarcompose2, funcs)
#kstarcompose = partial(reduce, kstarcompose2)
#use str.endswith( (tuple, of, vals)
extension = compose(itemgetter(-1), psplit('.'))
fileext = compose(extension, attr('filename'))
def iter_until_stop(f, *args, **kwargs):
while True:
try:
yield f(*args, **kwargs)
except StopIteration:
break
flatten_list = lambda a: a if type(a) != list else a[0]
def split_list(A, idx):
return A[:idx], A[idx:]
def ordered(articles):
return sorted(articles, key=attr('citation_count'), reverse=True)
def issues(infolist):
return filter(attr('cookies'), infolist)
def dump_path(path):
p = ", ".join(map(attr("__name__"), path))
return "{path} ==> {end}".format(path=p, end=lead_to(path))
def __str__(self):
path = ", ".join(map(attr("__name__"), self))
return "{path} ==> {end}".format(path=path, end=self.lead_to)
def field_names(self):
yield from map(attr('name'), fields(self.dataclass))
import functools
fsum(functools.partial(operator.mul, 2))(1,10)
############################################################
## currying (standard library)
############################################################
from operator import itemgetter
itemgetter(3)([1,3,5,7,9])
from operator import attrgetter as attr
class Speaker(object):
def __init__(self, name):
self.name = "[name] " + name
alexey = Speaker("alexey")
attr("name")(alexey)
from operator import methodcaller
methodcaller("__str__")([1,2,3,4,5])
methodcaller("keys")(dict(name="alexey", topic='fp'))
values_extractor = methodcaller("values")
values_extractor(dict(name="alexey", topic="fp"))
methodcaller("count", 1)([1,1,1,2,2])
############################################################
## good function is small function
############################################################
ss = ["UA", "PyCon", "2012"]
reduce(lambda acc, s: acc + len(s), ss, 0) ## BAD
reduce(lambda l,r: l+r, map(lambda s: len(s), ss)) ## NOT BAD...
reduce(operator.add, map(len, ss)) # GOOD
def seqs_equal(self, fn1, fn2, format):
open_sorted = compose(partial(sorted, key=compose(str, attr('seq'))),
partial(SeqIO.parse, format=format))
fq1, fq2 = map(open_sorted, [fn1, fn2])
self.assertFalse(len(fq1) == 0)
map(self.assert_seq_recs_equal, fq1, fq2)
: def _iterator_comparator iterator, value
: return value in iterator
: end
: def select name=None, choices=None, selected=None, **attributes
: """A factory for HTML <select> elements."""
: if isinstance(selected, (tuple, list, Iterator))
: comparator = _iterator_comparator
: else
: comparator = _basic_comparator
: end
<select&{attributes, name=name}>
: for group, options in groupby(options(choices), attr('group'))
: if group
<optgroup&{label=group.label, disabled=group.disabled}>
: end
: for option in options
<option&{value=option.value, disabled=option.disabled, selected=comparator(selected, option.value)}>${option.label}</option>
: end
: if group
</optgroup>
: end
: end
</select>
def main() -> None:
bots = list(get_data(today, [('func', Nanobot)]))
strongest = max(bots, key=attr('r'))
print(f'{today} star 1 = {sum(strongest.in_range(bot) for bot in bots)}')
print(f'{today} star 2 = {find_optimal_position(bots)}')
@urlpatterns.route("leaderboard/conversations/",
staff=True,
template="ej_gamification/leaderboard.jinja2")
def leaderboard_conversations():
scores = models.ConversationProgress.objects.filter(
score__gt=0).order_by("-score")
return {"leaderboard": prepare_dataframe(scores, FN_EXT, class_="table")}
#
# Constants and auxiliary functions
#
FN_BASE = {
_("Score"): attr("score"),
_("Votes"): attr("n_final_votes"),
_("Comments"): attr("n_approved_comments"),
_("Rejected"): attr("n_rejected_comments"),
_("pts"): lambda x: x.pts_approved_comments - x.pts_rejected_comments,
}
FN_EXT = {
_("Name"): attr("conversation.title"),
_("Author"): attr("conversation.author.name"),
**FN_BASE,
_("Conversation"): attr("total_conversation_score"),
}
def prepare_dataframe(data, transforms, **kwargs):
fns = transforms.values()
def dump_path(path):
p = ", ".join(map(attr("__name__"), path))
return "{path} ==> {end}".format(path = p, end = lead_to(path))
def __str__(self):
path = ", ".join(map(attr("__name__"), self))
return "{path} ==> {end}".format(path = path, end = self.lead_to)
### standard library currying
from operator import itemgetter
print(itemgetter(3)([1, 2, 3, 4, 5]))
from operator import attrgetter as attr
class Speaker(object):
def __init__(self, name):
self.name = "[name] " + name
alexey = Speaker("Alexey")
print(attr("name")(alexey))
from operator import methodcaller
methodcaller("__str__")([1, 2, 3, 4, 5])
print(methodcaller("keys")(dict(name="Alexey", topic="FP")))
values_extractor = methodcaller("values")
print(values_extractor(dict(name="Alexey", topic="FP")))
# good function is small function
ss = ["UA", "PyCon", "2012"]
#bad
reduce(lambda acc, s: acc + len(s), ss, 0)
# not bad
def fake_dirs(self):
return set(chain(*map(attr("parents"), self.fake_files)))
def plused_entries(self):
from operator import attrgetter as attr
return [p.entry for p in sorted(self.plused, key=attr('ctime'), reverse=True)]