def approximate(xs, ys, power):
assert len(xs) == len(ys)
matrix_size = power + 1
variables = 2 * power + 1
xs = map(float, xs)
ys = map(float, ys)
xs = reduce(
lambda x, y: x + [list(starmap(operator.mul, izip(x[-1], y)))], repeat(xs, variables - 1), [[1] * len(xs)]
)
assert len(xs) == variables
s = map(sum, xs)
assert s[0] == len(ys)
b = [sum(starmap(operator.mul, izip(ys, x))) for x in xs[:matrix_size]]
a = [s[i : i + matrix_size] for i in xrange(matrix_size)]
# So, we have a*x = b and we are looking for x
matr = [ai + [bi] for ai, bi in izip(a, b)]
def unify_row(i, j):
matr[i] = [cell / matr[i][j] for cell in matr[i]]
assert matr[i][j] == 1
def subtract_row(i, j, row):
assert matr[i][j] == 1
matr[row] = [matr[row][k] - matr[i][k] * matr[row][j] for k in xrange(len(matr[i]))]
assert matr[row][j] == 0
# NOTE: Example for matrix_size = 3
# unify_row(0, 0)
# subtract_row(0, 0, 1)
# subtract_row(0, 0, 2)
# unify_row(1, 1)
# subtract_row(1, 1, 2)
# unify_row(2, 2)
# subtract_row(2, 2, 1)
# subtract_row(2, 2, 0)
# subtract_row(1, 1, 0)
for i in xrange(matrix_size):
unify_row(i, i)
for j in xrange(matrix_size - i - 1):
subtract_row(i, i, i + j + 1)
for i in xrange(matrix_size):
for j in xrange(matrix_size - i - 1):
subtract_row(matrix_size - i - 1, matrix_size - i - 1, j)
assert all(matr[i][:matrix_size] == ([0] * i) + [1] + ([0] * (matrix_size - 1 - i)) for i in xrange(matrix_size))
ret = map(operator.itemgetter(matrix_size), matr)
return ret
def approximate(xs, ys, power):
assert len(xs) == len(ys)
matrix_size = power + 1
variables = 2 * power + 1
xs = map(float, xs)
ys = map(float, ys)
xs = reduce(lambda x, y: x + [list(starmap(operator.mul, izip(x[-1], y)))], repeat(xs, variables - 1), [[1] * len(xs)])
assert len(xs) == variables
s = map(sum, xs)
assert s[0] == len(ys)
b = [sum(starmap(operator.mul, izip(ys, x))) for x in xs[:matrix_size]]
a = [s[i:i + matrix_size] for i in xrange(matrix_size)]
# So, we have a*x = b and we are looking for x
matr = [ai + [bi] for ai, bi in izip(a, b)]
def unify_row(i, j):
matr[i] = [cell / matr[i][j] for cell in matr[i]]
assert matr[i][j] == 1
def subtract_row(i, j, row):
assert matr[i][j] == 1
matr[row] = [matr[row][k] - matr[i][k] * matr[row][j] for k in xrange(len(matr[i]))]
assert matr[row][j] == 0
# NOTE: Example for matrix_size = 3
# unify_row(0, 0)
# subtract_row(0, 0, 1)
# subtract_row(0, 0, 2)
# unify_row(1, 1)
# subtract_row(1, 1, 2)
# unify_row(2, 2)
# subtract_row(2, 2, 1)
# subtract_row(2, 2, 0)
# subtract_row(1, 1, 0)
for i in xrange(matrix_size):
unify_row(i, i)
for j in xrange(matrix_size - i - 1):
subtract_row(i, i, i + j + 1)
for i in xrange(matrix_size):
for j in xrange(matrix_size - i - 1):
subtract_row(matrix_size - i - 1, matrix_size - i - 1, j)
assert all(matr[i][:matrix_size] == ([0] * i) + [1] + ([0] * (matrix_size - 1 - i)) for i in xrange(matrix_size))
ret = map(operator.itemgetter(matrix_size), matr)
return ret
def __init__(self, personid_records):
'''
@param personid_records:
A list of tuples: (personid, bibrefrec, flag).
Notice that all bibrefrecs should be the same
since the Blob represents only one bibrefrec.
'''
self.bib = personid_records[0][1]
assert all(p[1] == self.bib for p in personid_records), \
"All cluster sets should share the bibrefrec"
self.claimed = set()
self.assigned = set()
self.rejected = set()
for pid, _, flag in personid_records:
if flag > 1:
self.claimed.add(pid)
elif flag >= -1:
self.assigned.add(pid)
else:
self.rejected.add(pid)
def __init__(self, personid_records):
'''
@param personid_records:
A list of tuples: (personid, bibrefrec, flag).
Notice that all bibrefrecs should be the same
since the Blob represents only one bibrefrec.
'''
self.bib = personid_records[0][1]
assert all(p[1] == self.bib for p in personid_records), \
"All cluster sets should share the bibrefrec"
self.claimed = set()
self.assigned = set()
self.rejected = set()
for pid, _, flag in personid_records:
if flag > 1:
self.claimed.add(pid)
elif flag >= -1:
self.assigned.add(pid)
else:
self.rejected.add(pid)
pid = os.fork()
if pid == 0: # child
os.nice(int((float(sizs[idx]) * 20.0 / biggest)))
run_job(job_idx)
else: # parent
pid_2_idx[pid] = job_idx
assert free > sizs[job_idx]
free -= sizs[job_idx]
del free_idxs[idx]
else:
break
pid, status = os.wait()
assert pid in pid_2_idx
idx = pid_2_idx[pid]
freed = sizs[idx]
done += freed
ret_status[idx] = status
free += freed
del pid_2_idx[pid]
update_status(done / total, "%d / %d" % (len(jobs) - len(free_idxs) - len(pid_2_idx), len(jobs)))
update_status_final("%d / %d" % (len(jobs), len(jobs)))
assert is_eq(free, initial)
assert not pid_2_idx
assert not free_idxs
assert len(jobs) == len(sizs) == len(ret_status) == len(bibs)
assert all(stat != None for stat in ret_status)
return ret_status
os.nice(int((float(sizs[idx]) * 20.0 / biggest)))
run_job(job_idx)
else: # parent
pid_2_idx[pid] = job_idx
assert free > sizs[job_idx]
free -= sizs[job_idx]
del free_idxs[idx]
else:
break
pid, status = os.wait()
assert pid in pid_2_idx
idx = pid_2_idx[pid]
freed = sizs[idx]
done += freed
ret_status[idx] = status
free += freed
del pid_2_idx[pid]
update_status(
done / total, "%d / %d" %
(len(jobs) - len(free_idxs) - len(pid_2_idx), len(jobs)))
update_status_final("%d / %d" % (len(jobs), len(jobs)))
assert is_eq(free, initial)
assert not pid_2_idx
assert not free_idxs
assert len(jobs) == len(sizs) == len(ret_status) == len(bibs)
assert all(stat != None for stat in ret_status)
return ret_status