def nanmin(a, axis=None):
"""Find the minimium over the given axis, ignoring NaNs.
"""
y = array(a,subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = _nx.inf
return y.min(axis)
def nanargmax(a, axis=None):
"""Find the maximum over the given axis ignoring NaNs.
"""
y = array(a,subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = -_nx.inf
return y.argmax(axis)
def nansum(a, axis=None):
"""Sum the array over the given axis, treating NaNs as 0.
"""
y = array(a, subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = 0
return y.sum(axis)
def nansum(a, axis=None):
"""Sum the array over the given axis, treating NaNs as 0.
"""
y = array(a,subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = 0
return y.sum(axis)
def nanmin(a, axis=None):
"""Find the minimium over the given axis, ignoring NaNs.
"""
y = array(a, subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = _nx.inf
return y.min(axis)
def nanargmax(a, axis=None):
"""Find the maximum over the given axis ignoring NaNs.
"""
y = array(a, subok=True)
if not issubclass(y.dtype.type, _nx.integer):
y[isnan(a)] = -_nx.inf
return y.argmax(axis)
def label_recall(confusion, label):
try:
recall = float(confusion[label, label]) / confusion[:, label].sum()
except:
recall = 0.0
return recall if not isnan(recall) else 0.0
def label_precision(confusion, label):
try:
precision = float(confusion[label, label]) / confusion[label, :].sum()
except:
precision = 0.0
return precision if not isnan(precision) else 0.0
def print_nonfinite(x):
with errstate(invalid='ignore'):
if umath.isnan(x):
ret = ('+' if sign == '+' else '') + nanstr
else: # isinf
infsgn = '-' if x < 0 else '+' if sign == '+' else ''
ret = infsgn + infstr
return ' ' * (pad_left + pad_right + 1 - len(ret)) + ret
def f1_macro(confusion, none_index, exclude_none=True):
p = precision_macro(confusion, none_index, exclude_none)
r = recall_macro(confusion, none_index, exclude_none)
try:
f1 = 2.0 * p * r / (p + r)
except:
f1 = 0.0
return f1 if not isnan(f1) else 0.0
def _clip_dep_is_scalar_nan(a):
# guarded to protect circular imports
from numpy.core.fromnumeric import ndim
if ndim(a) != 0:
return False
try:
return um.isnan(a)
except TypeError:
return False
def label_f1(confusion, label):
seterr(all='ignore')
p = label_precision(confusion, label)
r = label_recall(confusion, label)
try:
f1 = 2.0 * p * r / (p + r)
except:
f1 = 0.0
return f1 if not isnan(f1) else 0.0
def precision_micro(confusion, none_index, exclude_none=True):
n = confusion.shape[0]
num = sum([confusion[i, i] for i in range(n) if i != none_index or not exclude_none])
den = sum([confusion[i, :].sum() for i in range(n) if i != none_index or not exclude_none])
try:
precision = float(num) / den
except:
precision = 0.0
return precision if not isnan(precision) else 0.0
def recall_micro(confusion, none_index, exclude_none=True):
n = confusion.shape[0]
num = sum([confusion[i, i] for i in range(n) if i != none_index or not exclude_none])
den = sum([confusion[:, i].sum() for i in range(n) if i != none_index or not exclude_none])
try:
recall = float(num) / den
except:
recall = 0.0
return recall if not isnan(recall) else 0.0
def zsl_insert(self, score, ele):
"""
Insert a new node in the skiplist. Assumes the element does not already
exist (up to the caller to enforce that). The skiplist takes ownership
of the passed string 'ele'.
:param score:
:param ele:
:return:
"""
update = [None] * ZSKIPLIST_MAXLEVEL
rank = [0] * ZSKIPLIST_MAXLEVEL
assert (not isnan(score)), 'score can not be NaN'
x = self.header
i = self.level - 1
while i >= 0:
rank[i] = 0 if i == (self.level - 1) else rank[i + 1]
while (x.level[i].forward
and (x.level[i].forward.score < score or
(x.level[i].forward.score == score
and elecmp(x.level[i].forward.ele, ele) < 0))):
rank[i] += x.level[i].span
x = x.level[i].forward
update[i] = x
i -= 1
level = zsl_random_level()
if level > self.level:
for i in range(self.level, level):
rank[i] = 0
update[i] = self.header
update[i].level[i].span = self.length
self.level = level
x = ZskiplistNode(level=level, score=score, ele=ele)
for i in range(0, level):
x.level[i].forward = update[i].level[i].forward
update[i].level[i].forward = x
x.level[i].span = update[i].level[i].span - (rank[0] - rank[i])
update[i].level[i].span = (rank[0] - rank[i]) + 1
for i in range(level, self.level):
update[i].level[i].span += 1
x.backward = None if (update[0] is self.header) else update[0]
if x.level[0].forward:
x.level[0].forward.backward = x
else:
self.tail = x
self.length += 1
return x
def ecef_to_lat_lon_alt1(R, deg=True):
"""
Fukushima implementation of the Bowring algorithm (2006),
see [4] --
:param R: (X,Y,Z) -- coordinates in ECEF (numpy array)
:param deg: if True then lat and lon are returned in degrees (rad otherwise)
:return: (φ,θ,h) -- lat [deg], lon [deg], alt in WGS84 (numpy array)
"""
# WGS 84 constants
a = 6378137.0 # Equatorial Radius [m]
b = 6356752.314245 # Polar Radius [m]
e = 0.08181919092890624 # e = sqrt(1-b²/a²)
E = e ** 2
e1 = sqrt(1 - e ** 2) # e' = sqrt(1 - e²)
if isinstance(R, list): R = np.array(R)
p = sqrt(R[0] ** 2 + R[1] ** 2) # (1) - sqrt(X² + Y²)
az = abs(R[2])
Z = e1 * az / a
P = p / a
S, C = Z or 1, e1 * P or 1 # (C8) - zero approximation
max_iter = 5
for i in range(max_iter):
A = sqrt(S ** 2 + C ** 2)
Cn = P * A ** 3 - E * C ** 3
Sn = Z * A ** 3 + E * S ** 3
delta = abs(Sn / Cn - S / C) * C / S
if isnan(delta):
return ecef_to_lat_lon_alt1(R)
if abs(delta) < 1e-10 or i == max_iter - 1:
break
S, C = Sn, Cn
theta = np.math.atan2(R[1], R[0])
Cc = e1 * Cn
phi = np.sign(R[2]) * np.math.atan2(Sn, Cc)
h = (p * Cc + az * Sn - b * sqrt(Sn ** 2 + Cn ** 2)) / sqrt(Cc ** 2 + Sn ** 2)
if deg:
out = np.array([np.degrees(phi), np.degrees(theta), h])
else:
out = np.array([phi, theta, h])
# if filter(isnan, out):
# return ecef_to_lat_lon_alt1(R)
return out
def ecef_to_lat_lon_alt1(R, deg=True):
"""
Fukushima implementation of the Bowring algorithm (2006),
see [4] --
:param R: (X,Y,Z) -- coordinates in ECEF (numpy array)
:param deg: if True then lat and lon are returned in degrees (rad otherwise)
:return: (φ,θ,h) -- lat [deg], lon [deg], alt in WGS84 (numpy array)
"""
# WGS 84 constants
a = 6378137.0 # Equatorial Radius [m]
b = 6356752.314245 # Polar Radius [m]
e = 0.08181919092890624 # e = sqrt(1-b²/a²)
E = e**2
e1 = sqrt(1 - e**2) # e' = sqrt(1 - e²)
if isinstance(R, list): R = np.array(R)
p = sqrt(R[0]**2 + R[1]**2) # (1) - sqrt(X² + Y²)
az = abs(R[2])
Z = e1 * az / a
P = p / a
S, C = Z or 1, e1 * P or 1 # (C8) - zero approximation
max_iter = 5
for i in range(max_iter):
A = sqrt(S**2 + C**2)
Cn = P * A**3 - E * C**3
Sn = Z * A**3 + E * S**3
delta = abs(Sn / Cn - S / C) * C / S
if isnan(delta):
return ecef_to_lat_lon_alt1(R)
if abs(delta) < 1e-10 or i == max_iter - 1:
break
S, C = Sn, Cn
theta = np.math.atan2(R[1], R[0])
Cc = e1 * Cn
phi = np.sign(R[2]) * np.math.atan2(Sn, Cc)
h = (p * Cc + az * Sn - b * sqrt(Sn**2 + Cn**2)) / sqrt(Cc**2 + Sn**2)
if deg:
out = np.array([np.degrees(phi), np.degrees(theta), h])
else:
out = np.array([phi, theta, h])
# if filter(isnan, out):
# return ecef_to_lat_lon_alt1(R)
return out
def isnan(x):
if isinstance(x, IDataDescriptor):
return _um.isnan(dd_as_py(x))
else:
return _um.isnan(x)
def isnan(x):
# hacks to remove when isnan/isinf are available for data descriptors
if isinstance(x, IDataDescriptor):
return _um.isnan(dd_as_py(x))
else:
return _um.isnan(x)
def get_format_func(self, elem, **options):
missing_opt = self.check_options(**options)
if missing_opt:
raise Exception("Missing options: {}".format(missing_opt))
floatmode = options['floatmode']
precision = None if floatmode == 'unique' else options['precision']
suppress_small = options['suppress_small']
sign = options['sign']
infstr = options['infstr']
nanstr = options['nanstr']
exp_format = False
pad_left, pad_right = 0, 0
# only the finite values are used to compute the number of digits
finite = umath.isfinite(elem)
finite_vals = elem[finite]
nonfinite_vals = elem[~finite]
# choose exponential mode based on the non-zero finite values:
abs_non_zero = umath.absolute(finite_vals[finite_vals != 0])
if len(abs_non_zero) != 0:
max_val = np.max(abs_non_zero)
min_val = np.min(abs_non_zero)
with np.errstate(over='ignore'): # division can overflow
if max_val >= 1.e8 or (not suppress_small and
(min_val < 0.0001
or max_val / min_val > 1000.)):
exp_format = True
# do a first pass of printing all the numbers, to determine sizes
if len(finite_vals) == 0:
trim, exp_size, unique = '.', -1, True
elif exp_format:
trim, unique = '.', True
if floatmode == 'fixed':
trim, unique = 'k', False
strs = (format_float_scientific(x,
precision=precision,
unique=unique,
trim=trim,
sign=sign == '+')
for x in finite_vals)
frac_strs, _, exp_strs = zip(*(s.partition('e') for s in strs))
int_part, frac_part = zip(*(s.split('.') for s in frac_strs))
exp_size = max(len(s) for s in exp_strs) - 1
trim = 'k'
precision = max(len(s) for s in frac_part)
# this should be only 1 or 2. Can be calculated from sign.
pad_left = max(len(s) for s in int_part)
# pad_right is only needed for nan length calculation
pad_right = exp_size + 2 + precision
unique = False
else:
trim, unique = '.', True
if floatmode == 'fixed':
trim, unique = 'k', False
strs = (format_float_positional(x,
precision=precision,
fractional=True,
unique=unique,
trim=trim,
sign=sign == '+')
for x in finite_vals)
int_part, frac_part = zip(*(s.split('.') for s in strs))
pad_left = max(len(s) for s in int_part)
pad_right = max(len(s) for s in frac_part)
exp_size = -1
if floatmode in ['fixed', 'maxprec_equal']:
precision = pad_right
unique = False
trim = 'k'
else:
unique = True
trim = '.'
# account for sign = ' ' by adding one to pad_left
if sign == ' ' and not any(np.signbit(finite_vals)):
pad_left += 1
# account for nan and inf in pad_left
if len(nonfinite_vals) != 0:
nanlen, inflen = 0, 0
if np.any(umath.isinf(nonfinite_vals)):
neginf = sign != '-' or np.any(np.isneginf(nonfinite_vals))
inflen = len(infstr) + neginf
if np.any(umath.isnan(elem)):
nanlen = len(nanstr)
offset = pad_right + 1 # +1 for decimal pt
pad_left = max(nanlen - offset, inflen - offset, pad_left)
def print_nonfinite(x):
with errstate(invalid='ignore'):
if umath.isnan(x):
ret = ('+' if sign == '+' else '') + nanstr
else: # isinf
infsgn = '-' if x < 0 else '+' if sign == '+' else ''
ret = infsgn + infstr
return ' ' * (pad_left + pad_right + 1 - len(ret)) + ret
if exp_format:
def print_finite(x):
return format_float_scientific(x,
precision=precision,
unique=unique,
trim=trim,
sign=sign == '+',
pad_left=pad_left,
exp_digits=exp_size)
else:
def print_finite(x):
return format_float_positional(x,
precision=precision,
unique=unique,
fractional=True,
trim=trim,
sign=sign == '+',
pad_left=pad_left,
pad_right=pad_right)
def fmt(x):
if umath.isfinite(x):
return print_finite(x)
else:
return print_nonfinite(x)
return fmt
def zset_add(self, score, ele, flags):
"""
Add a new element or update the score of an existing element in a sorted
set.
The set of flags change the command behavior. They are passed with an integer
and will return other flags to indicate different conditions.
The input flags are the following:
ZADD_INCR: Increment the current element score by 'score' instead of updating
the current element score. If the element does not exist, we
assume 0 as previous score.
ZADD_NX: Perform the operation only if the element does not exist.
When ZADD_INCR is used, the new score of the element is returned with result.
The returned flags are the following:
ZADD_NAN: The resulting score is not a number.
ZADD_ADDED: The element was added (not present before the call).
ZADD_UPDATED: The element score was updated.
ZADD_NOP: No operation was performed because of NX.
Return value:
The function returns 1 on success, and sets the appropriate flags
ADDED or UPDATED to signal what happened during the operation (note that
none could be set if we re-added an element using the same score it used
to have, or in the case a zero increment is used).
The function returns 0 on erorr, currently only when the increment
produces a NAN condition, or when the 'score' value is NAN since the
start.
:param score:
:param ele:
:param flags:
:return:
"""
incr = (flags & ZADD_INCR) != 0
nx = (flags & ZADD_NX) != 0
if isnan(score):
return 0, ZADD_NAN, 0
dict_ = self.dict_
de = dict_.get(ele, None)
if de is not None:
if nx:
return 1, ZADD_NOP, 0
curscore = de
if incr:
score += curscore
if isnan(score):
return 0, ZADD_NAN, 0
# Remove and re-insert when score changes.
if score != curscore:
result = self.zsl.zsl_delete(score=curscore, ele=ele)
assert result == 1
result = self.zsl.zsl_insert(score=score, ele=ele)
assert result == 1
dict_[ele] = score
return 1, ZADD_UPDATED, score
else:
result = self.zsl.zsl_insert(score=score, ele=ele)
assert result == 1
assert dict_.setdefault(ele, score) == score
return 1, ZADD_ADDED, score
def isnan(x):
# hacks to remove when isnan/isinf are available for data descriptors
if isinstance(x, IDataDescriptor):
return _um.isnan(dd_as_py(x))
else:
return _um.isnan(x)
def tm_ransac5rows(d, sys):
class solstruct():
pass
sol = solstruct()
maxnrinl = 0
for iii in range(0, sys.ransac_k):
d2 = d ** 2
inl = (isfinite(d2)).astype(int)
r_c = d2.shape
m = r_c[0]
tmprows = random.permutation(m)
tmprows = tmprows[0:5]
auxvar1 = inl[tmprows, :]
auxvar2 = ((np.all(auxvar1, axis=0)).astype(int)).T
okcol = (np.flatnonzero(auxvar2)).T
B = d2[np.ix_(tmprows, okcol)]
ntmp = B.shape[1]
tmp2 = random.permutation(ntmp)
if ntmp > 5:
tmp21tup = tmp2[0:4]
tmp21 = np.reshape(tmp21tup, (1, -1))
tmp22tup = tmp2[4:, ]
tmp22 = np.reshape(tmp22tup, (1, -1))
cl, _ = compactionmatrix(5)
cr, _ = compactionmatrix(tmp2.shape[0])
Btmp1 = np.dot(cl, B[:, tmp2])
Btmp = np.dot(Btmp1, cr.conj().T)
B1 = Btmp[:, 0:3]
B2 = Btmp[:, 3:]
u, s, v = linalg.svd(B1)
u4tup = u[:, 3]
u4 = np.reshape(u4tup, (-1, 1))
if 0:
abs((u4.conj().T) * B2)
Imiss = isnan(d)
auxvar3 = abs(np.dot((u4.conj().T), B2))
okindtup = (auxvar3 > sys.ransac_threshold).nonzero()
okindmat = np.asarray(okindtup)
okind = np.reshape(okindmat, (1, -1))
inlim = zeros(d.shape)
inlim = inlim - Imiss
tmpconcat = concatenate((tmp21, tmp22[0, okind - 1]), 1)
tmprows = np.reshape(tmprows, (-1, 1))
inlim[tmprows, okcol[tmpconcat]] = ones((5, 4 + okind.size))
nrinl = 4 + okind.size
if nrinl > maxnrinl:
maxnrinl = nrinl
sol.rows = tmprows
concatmat = concatenate((tmp21, tmp22[0, okind - 1]), 1)
sol.cols = okcol[(concatmat)]
sol.row1 = sol.rows[1]
sol.col1 = sol.cols[0, 0]
sol.inlmatrix = inlim
B = d2[sol.rows, sol.cols]
cl, dl = compactionmatrix(B.shape[0])
cr, dr = compactionmatrix(B.shape[1])
Bhatdotprod = np.dot(dl, B)
Bhat = np.dot(Bhatdotprod, dr.conj().T)
Btildedotprod = np.dot(cl, B)
Btilde = np.dot(Btildedotprod, cr.conj().T)
u, s, vh = linalg.svd(Btilde)
v = vh.T
s[3:, ] = zeros(s.shape[0] - 3, s.shape[1])
Btilde = u * s * (v.conj().T)
Bhat[1:, 1:] = Btilde
sol.Bhat = Bhat
sol.dl = dl
sol.dr = dr
return sol, maxnrinl
