def __div__(self, amount):
if isinstance(amount, Duration) and amount.month:
m = self.month
r = self.milli
# DO NOT CONSIDER TIME OF DAY
tod = r % MILLI_VALUES.day
r = r - tod
if m == 0 and r > (MILLI_VALUES.year / 3):
m = Math.floor(12 * self.milli / MILLI_VALUES.year)
r -= (m / 12) * MILLI_VALUES.year
else:
r = r - (self.month * MILLI_VALUES.month)
if r >= MILLI_VALUES.day * 31:
from pyLibrary.debugs.logs import Log
Log.error("Do not know how to handle")
r = MIN(29 / 30, (r + tod) / (MILLI_VALUES.day * 30))
output = Math.floor(m / amount.month) + r
return output
elif Math.is_number(amount):
output = Duration(0)
output.milli = self.milli / amount
output.month = self.month / amount
return output
else:
return self.milli / amount.milli
def __new__(cls, value=None, **kwargs):
output = object.__new__(cls)
if value == None:
if kwargs:
output.milli = datetime.timedelta(**kwargs).total_seconds() * 1000
output.month = 0
return output
else:
return None
if Math.is_number(value):
output._milli = float(value) * 1000
output.month = 0
return output
elif isinstance(value, basestring):
return parse(value)
elif isinstance(value, Duration):
output.milli = value.milli
output.month = value.month
return output
elif isinstance(value, float) and Math.is_nan(value):
return None
else:
from pyLibrary import convert
from pyLibrary.debugs.logs import Log
Log.error("Do not know type of object (" + convert.value2json(value) + ")of to make a Duration")
def __init__(self, **desc):
Domain.__init__(self, **desc)
self.type = "range"
self.NULL = Null
if self.partitions:
# IGNORE THE min, max, interval
if not self.key:
Log.error("Must have a key value")
parts = listwrap(self.partitions)
for i, p in enumerate(parts):
self.min = Math.min(self.min, p.min)
self.max = Math.max(self.max, p.max)
if p.dataIndex != None and p.dataIndex != i:
Log.error("Expecting `dataIndex` to agree with the order of the parts")
if p[self.key] == None:
Log.error("Expecting all parts to have {{key}} as a property", key=self.key)
p.dataIndex = i
# VERIFY PARTITIONS DO NOT OVERLAP, HOLES ARE FINE
for p, q in itertools.product(parts, parts):
if p.min <= q.min and q.min < p.max:
Log.error("partitions overlap!")
self.partitions = parts
return
elif any([self.min == None, self.max == None, self.interval == None]):
Log.error("Can not handle missing parameter")
self.key = "min"
self.partitions = wrap([{"min": v, "max": v + self.interval, "dataIndex": i} for i, v in enumerate(frange(self.min, self.max, self.interval))])
def geo_mean(values):
"""
GIVEN AN ARRAY OF dicts, CALC THE GEO-MEAN ON EACH ATTRIBUTE
"""
agg = Struct()
for d in values:
for k, v in d.items():
if v != 0:
agg[k] = nvl(agg[k], ZeroMoment.new_instance()) + Math.log(Math.abs(v))
return {k: Math.exp(v.stats.mean) for k, v in agg.items()}
def test_floor_mod_identity_w_ints(self):
for i in range(100):
x = Random.float()*200 - 100.0
m = floor(abs(random.gauss(0, 5)))
if m == 0:
self.assertEqual(Math.floor(x, m), None)
self.assertEqual(Math.mod(x, m), None)
else:
self.assertAlmostEqual(Math.floor(x, m)+Math.mod(x, m), x, places=7)
def assertAlmostEqualValue(test, expected, digits=None, places=None, msg=None, delta=None):
"""
Snagged from unittest/case.py, then modified (Aug2014)
"""
if expected == None: # None has no expectations
return
if test == expected:
# shortcut
return
if not Math.is_number(expected):
# SOME SPECIAL CASES, EXPECTING EMPTY CONTAINERS IS THE SAME AS EXPECTING NULL
if isinstance(expected, list) and len(expected)==0 and test == None:
return
if isinstance(expected, Mapping) and not expected.keys() and test == None:
return
if test != expected:
raise AssertionError(expand_template("{{test}} != {{expected}}", locals()))
return
num_param = 0
if digits != None:
num_param += 1
if places != None:
num_param += 1
if delta != None:
num_param += 1
if num_param>1:
raise TypeError("specify only one of digits, places or delta")
if digits is not None:
with suppress_exception:
diff = Math.log10(abs(test-expected))
if diff < digits:
return
standardMsg = expand_template("{{test}} != {{expected}} within {{digits}} decimal places", locals())
elif delta is not None:
if abs(test - expected) <= delta:
return
standardMsg = expand_template("{{test}} != {{expected}} within {{delta}} delta", locals())
else:
if places is None:
places = 15
with suppress_exception:
diff = Math.log10(abs(test-expected))
if diff < Math.ceiling(Math.log10(abs(test)))-places:
return
standardMsg = expand_template("{{test|json}} != {{expected|json}} within {{places}} places", locals())
raise AssertionError(coalesce(msg, "") + ": (" + standardMsg + ")")
def intervals(_min, _max=None, size=1):
"""
RETURN (min, max) PAIRS OF GIVEN SIZE, WHICH COVER THE _min, _max RANGE
THE LAST PAIR MAY BE SMALLER
Yes! It's just like range(), only cooler!
"""
if _max == None:
_max = _min
_min = 0
_max = int(Math.ceiling(_max))
_min = int(Math.floor(_min))
output = ((x, min(x + size, _max)) for x in __builtin__.range(_min, _max, size))
return output
def intervals(_min, _max=None, size=1):
"""
RETURN (min, max) PAIRS OF GIVEN SIZE, WHICH COVER THE _min, _max RANGE
THE LAST PAIR MAY BE SMALLER
Yes! It's just like range(), only cooler!
"""
if _max == None:
_max = _min
_min = 0
_max = int(Math.ceiling(_max))
_min = int(Math.floor(_min))
output = ((x, min(x + size, _max)) for x in __builtin__.range(_min, _max, size))
return output
def ighmm_log_gamma_sum(log_a, s, parent):
max = 1.0
argmax = 0
# shortcut for the trivial case
if parent.gamma_states == 1:
return parent.gamma_a[0] + log_a[parent.gamma_id[0]]
logP = ARRAY_MALLOC(len(s.in_a))
# calculate logs of a[k,l]*gamma[k,hi] as sums of logs and find maximum:
for j in range(len(s.in_a)):
# search for state j_id in the gamma list
for k in range(0, parent.gamma_states):
if parent.gamma_id[k] == j:
break
if k == parent.gamma_states:
logP[j] = 1.0
else:
logP[j] = log_a[j] + parent.gamma_a[k]
if max == 1.0 or (logP[j] > max and logP[j] != 1.0):
max = logP[j]
argmax = j
# calculate max+Math.log(1+sum[j!=argmax exp(logP[j]-max)])
result = 1.0
for j in range(len(s.in_a)):
if j != argmax and logP[j] != 1.0:
result += exp(logP[j] - max)
result = Math.log(result)
result += max
return result
def pop(self, wait=SECOND, till=None):
m = self.queue.read(wait_time_seconds=Math.floor(wait.seconds))
if not m:
return None
self.pending.append(m)
return convert.json2value(m.get_body())
def pdf(self, data):
# XXX assume root as first index
assert self.parents[0] == -1
assert self.w[0] == 0.0
res = np.zeros(len(data))
for i in range(len(data)):
res[i] = Math.log((1.0 / (math.sqrt(2.0 * math.pi) * self.variance[0])) * math.exp(( data[i, 0] - self.mean[0] ) ** 2 / (-2.0 * self.variance[0] ** 2)))
for j in range(1, self.dimension):
pind = self.parents[j]
res[i] += Math.log(
(1.0 / (math.sqrt(2.0 * math.pi) * self.variance[j])) * math.exp(( data[i, j] - self.mean[j] - self.w[j] * ( data[i, pind] - self.mean[pind] ) ) ** 2 / (-2.0 * self.variance[j] ** 2)))
return res
def convert(self, expr):
"""
ADD THE ".$value" SUFFIX TO ALL VARIABLES
"""
if expr is True or expr == None or expr is False:
return expr
elif Math.is_number(expr):
return expr
elif expr == ".":
return "."
elif is_keyword(expr):
#TODO: LOOKUP SCHEMA AND ADD ALL COLUMNS WITH THIS PREFIX
return expr + ".$value"
elif isinstance(expr, basestring):
Log.error("{{name|quote}} is not a valid variable name", name=expr)
elif isinstance(expr, Date):
return expr
elif isinstance(expr, Query):
return self._convert_query(expr)
elif isinstance(expr, Mapping):
if expr["from"]:
return self._convert_query(expr)
elif len(expr) >= 2:
#ASSUME WE HAVE A NAMED STRUCTURE, NOT AN EXPRESSION
return wrap({name: self.convert(value) for name, value in expr.items()})
else:
# ASSUME SINGLE-CLAUSE EXPRESSION
k, v = expr.items()[0]
return self.converter_map.get(k, self._convert_bop)(k, v)
elif isinstance(expr, (list, set, tuple)):
return wrap([self.convert(value) for value in expr])
def sample(self, native=False):
if native:
return random.normalvariate(self.mean, self.variance)
else:
r2 = -2.0 * Math.log(random_mt.float23()) # r2 ~ chi-square(2)
theta = 2.0 * math.pi * random_mt.float23() # theta ~ uniform(0, 2 \pi)
return math.sqrt(self.variance) * math.sqrt(r2) * math.cos(theta) + self.mean
def wrap_gsl_dirichlet_lnpdf(alpha, x):
if hasattr(x[0], "__iter__"):
output = [wrap_gsl_dirichlet_lnpdf(alpha, xi) for xi in x]
else:
output = Math.log(special.gamma(sum(alpha))) - np.sum(np.log(special.gamma(alpha))) + np.sum(np.log([xi ** (ai - 1.0) for xi, ai in zip(x, alpha)]))
return output
def parse(*args):
try:
if len(args) == 1:
a0 = args[0]
if isinstance(a0, (datetime, date)):
output = unix2Date(datetime2unix(a0))
elif isinstance(a0, Date):
output = unix2Date(a0.unix)
elif isinstance(a0, (int, long, float, Decimal)):
a0 = float(a0)
if a0 > 9999999999: # WAY TOO BIG IF IT WAS A UNIX TIMESTAMP
output = unix2Date(a0 / 1000)
else:
output = unix2Date(a0)
elif isinstance(a0, basestring) and len(a0) in [9, 10, 12, 13] and Math.is_integer(a0):
a0 = float(a0)
if a0 > 9999999999: # WAY TOO BIG IF IT WAS A UNIX TIMESTAMP
output = unix2Date(a0 / 1000)
else:
output = unix2Date(a0)
elif isinstance(a0, basestring):
output = unicode2Date(a0)
else:
output = unix2Date(datetime2unix(datetime(*args)))
else:
if isinstance(args[0], basestring):
output = unicode2Date(*args)
else:
output = unix2Date(datetime2unix(datetime(*args)))
return output
except Exception, e:
from pyLibrary.debugs.logs import Log
Log.error("Can not convert {{args}} to Date", args=args, cause=e)
def end(self):
ignore = Math.ceiling(len(self.samples) * (1 - self.middle) / 2)
if ignore * 2 >= len(self.samples):
return stats.Stats()
output = stats.Stats(samples=sorted(self.samples)[ignore : len(self.samples) - ignore :])
output.samples = list(self.samples)
return output
def __init__(self, edge, query, limit):
AggsDecoder.__init__(self, edge, query, limit)
self.domain = edge.domain
self.domain.limit =Math.min(coalesce(self.domain.limit, query.limit, 10), MAX_LIMIT)
self.parts = list()
self.key2index = {}
self.computed_domain = False
def quote_value(self, value):
"""
convert values to mysql code for the same
mostly delegate directly to the mysql lib, but some exceptions exist
"""
try:
if value == None:
return "NULL"
elif isinstance(value, SQL):
if not value.param:
# value.template CAN BE MORE THAN A TEMPLATE STRING
return self.quote_sql(value.template)
param = {k: self.quote_sql(v) for k, v in value.param.items()}
return expand_template(value.template, param)
elif isinstance(value, basestring):
return self.db.literal(value)
elif isinstance(value, datetime):
return "str_to_date('" + value.strftime("%Y%m%d%H%M%S") + "', '%Y%m%d%H%i%s')"
elif hasattr(value, '__iter__'):
return self.db.literal(json_encode(value))
elif isinstance(value, Mapping):
return self.db.literal(json_encode(value))
elif Math.is_number(value):
return unicode(value)
else:
return self.db.literal(value)
except Exception, e:
Log.error("problem quoting SQL", e)
def convert(self, expr):
"""
EXPAND INSTANCES OF name TO value
"""
if expr is True or expr == None or expr is False:
return expr
elif Math.is_number(expr):
return expr
elif expr == ".":
return "."
elif is_keyword(expr):
return coalesce(self.dimensions[expr], expr)
elif isinstance(expr, basestring):
Log.error("{{name|quote}} is not a valid variable name", name=expr)
elif isinstance(expr, Date):
return expr
elif isinstance(expr, Query):
return self._convert_query(expr)
elif isinstance(expr, Mapping):
if expr["from"]:
return self._convert_query(expr)
elif len(expr) >= 2:
#ASSUME WE HAVE A NAMED STRUCTURE, NOT AN EXPRESSION
return wrap({name: self.convert(value) for name, value in expr.leaves()})
else:
# ASSUME SINGLE-CLAUSE EXPRESSION
k, v = expr.items()[0]
return converter_map.get(k, self._convert_bop)(self, k, v)
elif isinstance(expr, (list, set, tuple)):
return wrap([self.convert(value) for value in expr])
else:
return expr
def icompressed2ibytes(source):
"""
:param source: GENERATOR OF COMPRESSED BYTES
:return: GENERATOR OF BYTES
"""
decompressor = zlib.decompressobj(16 + zlib.MAX_WBITS)
last_bytes_count = 0 # Track the last byte count, so we do not show too many debug lines
bytes_count = 0
for bytes_ in source:
data = decompressor.decompress(bytes_)
bytes_count += len(data)
if Math.floor(last_bytes_count, 1000000) != Math.floor(bytes_count, 1000000):
last_bytes_count = bytes_count
if DEBUG:
Log.note("bytes={{bytes}}", bytes=bytes_count)
yield data
def int2Partition(value):
if Math.round(value) == 0:
return edge.domain.NULL
d = datetime(str(value)[:4:], str(value)[-2:], 1)
d = d.addMilli(offset)
return edge.domain.getPartByKey(d)
def test_mod(self):
self.assertEqual(Math.mod(12, 12), 0)
self.assertEqual(Math.mod(11, 12), 11)
self.assertEqual(Math.mod(2, 12), 2)
self.assertEqual(Math.mod(1, 12), 1)
self.assertEqual(Math.mod(-0, 12), 0)
self.assertEqual(Math.mod(-1, 12), 11)
self.assertEqual(Math.mod(-2, 12), 10)
self.assertEqual(Math.mod(-12, 12), 0)
def get_all_vars(expr):
if expr == None:
return set()
elif isinstance(expr, unicode):
if expr == "." or is_keyword(expr):
return set([expr])
else:
Log.error("Expecting a json path")
elif expr is True:
return set()
elif expr is False:
return set()
elif Math.is_number(expr):
return set()
op, term = expr.items()[0]
mop = ruby_multi_operators.get(op)
if mop:
if isinstance(term, list):
output = set()
for t in term:
output |= get_all_vars(t)
return output
elif isinstance(term, Mapping):
a, b = term.items()[0]
return get_all_vars(a) | get_all_vars(b)
else:
get_all_vars(term)
bop = ruby_binary_operators.get(op)
if bop:
if isinstance(term, list):
output = set()
for t in term:
output |= get_all_vars(t)
return output
elif isinstance(term, Mapping):
if op == "eq":
output = set()
for a, b in term.items():
output |= get_all_vars(
a) # {k:v} k IS VARIABLE, v IS A VALUE
return output
else:
a, b = term.items()[0]
return get_all_vars(a)
else:
Log.error("Expecting binary term")
uop = ruby_unary_operators.get(op)
if uop:
return get_all_vars(term)
cop = complex_operators.get(op)
if cop:
return cop(op, term).vars()
Log.error("`{{op}}` is not a recognized operation", op=op)
def wrap(query, schema=None):
"""
NORMALIZE QUERY SO IT CAN STILL BE JSON
"""
if isinstance(query, QueryOp) or query == None:
return query
query = wrap(query)
output = QueryOp("from", None)
output.format = query.format
output.frum = wrap_from(query["from"], schema=schema)
if not schema and isinstance(output.frum, Schema):
schema = output.frum
if query.select:
output.select = _normalize_selects(query.select, query.frum, schema=schema)
else:
if query.edges or query.groupby:
output.select = Dict(name="count", value=jx_expression("."), aggregate="count", default=0)
else:
output.select = _normalize_selects(".", query["from"])
if query.groupby and query.edges:
Log.error("You can not use both the `groupby` and `edges` clauses in the same query!")
elif query.edges:
output.edges = _normalize_edges(query.edges, schema=schema)
output.groupby = Null
elif query.groupby:
output.edges = Null
output.groupby = _normalize_groupby(query.groupby, schema=schema)
else:
output.edges = Null
output.groupby = Null
output.where = _normalize_where(query.where, schema=schema)
output.window = [_normalize_window(w) for w in listwrap(query.window)]
output.having = None
output.sort = _normalize_sort(query.sort)
output.limit = Math.min(MAX_LIMIT, coalesce(query.limit, DEFAULT_LIMIT))
if not Math.is_integer(output.limit) or output.limit < 0:
Log.error("Expecting limit >= 0")
output.isLean = query.isLean
return output
def get_all_vars(expr):
if expr == None:
return set()
elif isinstance(expr, unicode):
if expr == "." or is_keyword(expr):
return set([expr])
else:
Log.error("Expecting a json path")
elif expr is True:
return set()
elif expr is False:
return set()
elif Math.is_number(expr):
return set()
op, term = expr.items()[0]
mop = ruby_multi_operators.get(op)
if mop:
if isinstance(term, list):
output = set()
for t in term:
output |= get_all_vars(t)
return output
elif isinstance(term, Mapping):
a, b = term.items()[0]
return get_all_vars(a) | get_all_vars(b)
else:
get_all_vars(term)
bop = ruby_binary_operators.get(op)
if bop:
if isinstance(term, list):
output = set()
for t in term:
output |= get_all_vars(t)
return output
elif isinstance(term, Mapping):
if op == "eq":
output = set()
for a, b in term.items():
output |= get_all_vars(a) # {k:v} k IS VARIABLE, v IS A VALUE
return output
else:
a, b = term.items()[0]
return get_all_vars(a)
else:
Log.error("Expecting binary term")
uop = ruby_unary_operators.get(op)
if uop:
return get_all_vars(term)
cop = complex_operators.get(op)
if cop:
return cop(op, term).vars()
Log.error("`{{op}}` is not a recognized operation", op= op)
def value2query(value):
if isinstance(value, datetime):
return convert.datetime2milli(value)
if isinstance(value, Duration):
return value.milli
if Math.is_number(value):
return value
return convert.string2quote(value)
def Viterbi_precompute(mo, o, len, v):
# Precomputing the Math.log(a_ij)
for j in range(mo.N):
for i in range(mo.N):
if mo.s[j].in_a[i] == 0.0: # DBL_EPSILON ?
v.log_in_a[j][i] = +1 # Not used any further in the calculations
else:
v.log_in_a[j][i] = Math.log(mo.s[j].in_a[i])
# Precomputing the Math.log(bj(ot))
for j in range(mo.N):
for t in range(mo.M):
if mo.s[j].b[t] == 0.0: # DBL_EPSILON ?
v.log_b[j][t] = +1
else:
v.log_b[j][t] = Math.log(mo.s[j].b[t])
def floor(self, interval=None):
if not isinstance(interval, Duration):
from pyLibrary.debugs.logs import Log
Log.error("Expecting an interval as a Duration object")
output = Duration(0)
if interval.month:
if self.month:
output.month = int(Math.floor(self.month / interval.month) * interval.month)
output.milli = output.month * MILLI_VALUES.month
return output
# A MONTH OF DURATION IS BIGGER THAN A CANONICAL MONTH
output.month = int(Math.floor(self.milli * 12 / MILLI_VALUES["year"] / interval.month) * interval.month)
output.milli = output.month * MILLI_VALUES.month
else:
output.milli = Math.floor(self.milli / (interval.milli)) * (interval.milli)
return output
def kbest_buildLogMatrix(s, N):
# create & initialize matrix:
log_a = ARRAY_MALLOC(N)
for i in range(0, N):
log_a[i] = ARRAY_MALLOC(N)
for j in range(N):
log_a[i][j] = Math.log(s[i].in_a[j])
return log_a
def find_keys(self, start, count, filter=None):
digits = int(Math.ceiling(log10(count - 1)))
prefix = unicode(start)[:-digits]
metas = self.bucket.metas(prefix=prefix)
min_ = Version(unicode(start))
max_ = Version(unicode(start+count))
output = [m.key for m in metas if min_ <= Version(m.key) < max_]
return set(output)
def sviterbi_precompute(smo, O, T, v):
# Precomputing of Math.log(b_j(O_t))
for t in range(T):
for j in range(smo.N):
cb = smo.s[j].calc_b(O[t])
if cb == 0.0:
# DBL_EPSILON ?
v.log_b[j][t] = -DBL_MAX
else:
v.log_b[j][t] = Math.log(cb)
def pop_message(self, wait=SECOND, till=None):
"""
RETURN THE MESSAGE, CALLER IS RESPONSIBLE FOR CALLING delete_message() WHEN DONE
"""
m = self.queue.read(wait_time_seconds=Math.floor(wait.seconds))
if not m:
return None
output = convert.json2value(m.get_body())
return output
