Python numpy 模块，ones_like() 实例源码
我们从Python开源项目中，提取了以下50个代码示例，用于说明如何使用numpy.ones_like()。
def blur(I, r):
"""
This method performs like cv2.blur().
Parameters
----------
I: NDArray
Filtering input
r: int
Radius of blur filter
Returns
-------
q: NDArray
Blurred output of I.
"""
ones = np.ones_like(I, dtype=np.float32)
N = box_filter(ones, r)
ret = box_filter(I, r)
return ret / N
def load_ROI_mask(self):
proxy = nib.load(self.FLAIR_FILE)
image_array = np.asarray(proxy.dataobj)
mask = np.ones_like(image_array)
mask[np.where(image_array < 90)] = 0
# img = nib.Nifti1Image(mask, proxy.affine)
# nib.save(img, join(modalities_path,'mask.nii.gz'))
struct_element_size = (20, 20, 20)
mask_augmented = np.pad(mask, [(21, 21), (21, 21), (21, 21)], 'constant', constant_values=(0, 0))
mask_augmented = binary_closing(mask_augmented, structure=np.ones(struct_element_size, dtype=bool)).astype(
np.int)
return mask_augmented[21:-21, 21:-21, 21:-21].astype('bool')
def make_3d_mask(img_shape, center, radius, shape='sphere'):
mask = np.zeros(img_shape)
radius = np.rint(radius)
center = np.rint(center)
sz = np.arange(int(max(center[0] - radius, 0)), int(max(min(center[0] + radius + 1, img_shape[0]), 0)))
sy = np.arange(int(max(center[1] - radius, 0)), int(max(min(center[1] + radius + 1, img_shape[1]), 0)))
sx = np.arange(int(max(center[2] - radius, 0)), int(max(min(center[2] + radius + 1, img_shape[2]), 0)))
sz, sy, sx = np.meshgrid(sz, sy, sx)
if shape == 'cube':
mask[sz, sy, sx] = 1.
elif shape == 'sphere':
distance2 = ((center[0] - sz) ** 2
+ (center[1] - sy) ** 2
+ (center[2] - sx) ** 2)
distance_matrix = np.ones_like(mask) * np.inf
distance_matrix[sz, sy, sx] = distance2
mask[(distance_matrix <= radius ** 2)] = 1
elif shape == 'gauss':
z, y, x = np.ogrid[:mask.shape[0], :mask.shape[1], :mask.shape[2]]
distance = ((z - center[0]) ** 2 + (y - center[1]) ** 2 + (x - center[2]) ** 2)
mask = np.exp(- 1. * distance / (2 * radius ** 2))
mask[(distance > 3 * radius ** 2)] = 0
return mask
def sim_target_fixed(target_data, target_labels, sigma, idx, target_params):
"""
Sets as target to have fixed similarity between all the training samples
:param target_data: (not used)
:param target_labels: (not used)
:param sigma: not used
:param idx: indices of the data samples to be used for the calculation of the similarity matrix
:param target_params: expect to found the 'target_value' here
:return: the similarity matrix and the corresponding mask
"""
if 'target_value' not in target_params:
target_params['target_value'] = 0.0
Gt = np.ones((len(idx), len(idx)))
Gt = Gt * target_params['target_value']
Gt_mask = np.ones_like(Gt)
return np.float32(Gt), np.float32(Gt_mask)
def getCylinderPoints(xc,zc,r):
xLocOrig1 = np.arange(-r,r+r/10.,r/10.)
xLocOrig2 = np.arange(r,-r-r/10.,-r/10.)
# Top half of cylinder
zLoc1 = np.sqrt(-xLocOrig1**2.+r**2.)+zc
# Bottom half of cylinder
zLoc2 = -np.sqrt(-xLocOrig2**2.+r**2.)+zc
# Shift from x = 0 to xc
xLoc1 = xLocOrig1 + xc*np.ones_like(xLocOrig1)
xLoc2 = xLocOrig2 + xc*np.ones_like(xLocOrig2)
topHalf = np.vstack([xLoc1,zLoc1]).T
topHalf = topHalf[0:-1,:]
bottomhalf = np.vstack([xLoc2,zLoc2]).T
bottomhalf = bottomhalf[0:-1,:]
cylinderPoints = np.vstack([topHalf,bottomhalf])
cylinderPoints = np.vstack([cylinderPoints,topHalf[0,:]])
return cylinderPoints
def test_basic(self):
dts = [np.bool, np.int16, np.int32, np.int64, np.double, np.complex128,
np.longdouble, np.clongdouble]
for dt in dts:
c = np.ones(53, dtype=np.bool)
assert_equal(np.where( c, dt(0), dt(1)), dt(0))
assert_equal(np.where(~c, dt(0), dt(1)), dt(1))
assert_equal(np.where(True, dt(0), dt(1)), dt(0))
assert_equal(np.where(False, dt(0), dt(1)), dt(1))
d = np.ones_like(c).astype(dt)
e = np.zeros_like(d)
r = d.astype(dt)
c[7] = False
r[7] = e[7]
assert_equal(np.where(c, e, e), e)
assert_equal(np.where(c, d, e), r)
assert_equal(np.where(c, d, e[0]), r)
assert_equal(np.where(c, d[0], e), r)
assert_equal(np.where(c[::2], d[::2], e[::2]), r[::2])
assert_equal(np.where(c[1::2], d[1::2], e[1::2]), r[1::2])
assert_equal(np.where(c[::3], d[::3], e[::3]), r[::3])
assert_equal(np.where(c[1::3], d[1::3], e[1::3]), r[1::3])
assert_equal(np.where(c[::-2], d[::-2], e[::-2]), r[::-2])
assert_equal(np.where(c[::-3], d[::-3], e[::-3]), r[::-3])
assert_equal(np.where(c[1::-3], d[1::-3], e[1::-3]), r[1::-3])
def plot_series(series):
plt.figure(1)
# colors = [np.array([1, 0.1, 0.1]), np.array([0.1, 1, 0.1]), np.array([0.1, 0.1, 1])]
colors = ['m', 'g', 'r', 'b', 'y']
for i, s in enumerate(series):
print(s['x'], s['y'], s['std'], s['label'])
small_number = np.ones_like(s['x']) * (s['x'][1]*0.1)
x_axis = np.where(s['x'] == 0, small_number, s['x'])
plt.plot(x_axis, s['y'], color=colors[i], label=s['label'])
plt.fill_between(x_axis, s['y'] - s['std'], s['y'] + s['std'], color=colors[i], alpha=0.2)
plt.semilogx()
plt.xlabel('MI reward bonus')
plt.ylabel('Final intrinsic reward')
plt.title('Final intrinsic reward in pointMDP with 10 good modes')
plt.legend(loc='best')
plt.show()
def predict(self, path):
if 'env_infos' in path.keys() and 'full_path' in path['env_infos'].keys():
expanded_path = tensor_utils.flatten_first_axis_tensor_dict(path['env_infos']['full_path'])
else: # when it comes from log_diagnostics it's already expanded (or if it was never aggregated)
expanded_path = path
bonus = self.visitation_bonus * self.predict_count(expanded_path) + \
self.dist_from_reset_bonus * self.predict_dist_from_reset(expanded_path)
if self.snn_H_bonus: # I need the if because the snn bonus is only available when there are latents
bonus += self.snn_H_bonus * self.predict_entropy(expanded_path)
total_bonus = bonus + self.survival_bonus * np.ones_like(bonus)
if 'env_infos' in path.keys() and 'full_path' in path['env_infos'].keys():
aggregated_bonus = []
full_path_rewards = path['env_infos']['full_path']['rewards']
total_steps = 0
for sub_rewards in full_path_rewards:
aggregated_bonus.append(np.sum(total_bonus[total_steps:total_steps + len(sub_rewards)]))
total_steps += len(sub_rewards)
total_bonus = aggregated_bonus
return np.array(total_bonus)
def _initialize_filter(self):
"""Set up spectral filter or dealiasing."""
if self.use_filter:
cphi=0.65*pi
wvx=np.sqrt((self.k*self.dx)**2.+(self.l*self.dy)**2.)
self.filtr = np.exp(-23.6*(wvx-cphi)**4.)
self.filtr[wvx<=cphi] = 1.
self.logger.info(' Using filter')
elif self.dealias:
self.filtr = np.ones_like(self.wv2)
self.filtr[self.nx/3:2*self.nx/3,:] = 0.
self.filtr[:,self.ny/3:2*self.ny/3] = 0.
self.logger.info(' Dealiasing with 2/3 rule')
else:
self.filtr = np.ones_like(self.wv2)
self.logger.info(' No dealiasing; no filter')
def _initialize_filter(self):
""" Set up spectral filter or dealiasing."""
if self.use_filter:
cphi=0.65*pi
wvx=np.sqrt((self.k*self.dx)**2.+(self.l*self.dy)**2.)
self.filtr = np.exp(-23.6*(wvx-cphi)**4.)
self.filtr[wvx<=cphi] = 1.
self.logger.info(' Using filter')
elif self.dealias:
self.filtr = np.ones_like(self.wv2)
self.filtr[self.nx//3:2*self.nx//3,:] = 0.
self.filtr[:,self.ny//3:2*self.ny//3] = 0.
self.logger.info(' Dealiasing with 2/3 rule')
else:
self.filtr = np.ones_like(self.wv2)
self.logger.info(' No dealiasing; no filter')
def flux(self, time, planets = None, cadence = 'lc'):
'''
'''
# Ensure it's a list
if planets is None:
planets = self.planets
elif type(planets) is str:
planets = [planets]
# Compute flux for each planet
flux = np.ones_like(time)
for planet in planets:
if cadence == 'lc':
model = ps.Transit(per = self.period[planet], b = self.b[planet], RpRs = self.RpRs[planet], t0 = self.t0[planet],
rhos = self.rhos, ecc = self.ecc[planet], w = self.w[planet] * np.pi / 180., u1 = self.u1,
u2 = self.u2, times = self.times[planet])
else:
model = ps.Transit(per = self.period[planet], b = self.b[planet], RpRs = self.RpRs[planet], t0 = self.t0[planet],
rhos = self.rhos, ecc = self.ecc[planet], w = self.w[planet] * np.pi / 180., u1 = self.u1,
u2 = self.u2, times = self.times[planet], exptime = ps.KEPSHRTCAD)
flux *= model(time)
return flux
def test_flat_tensor_dot_tensor():
"""
Ensure that a flattened argument axis is not unflattend in the result.
"""
H = ng.make_axis(2)
W = ng.make_axis(7)
C = ng.make_axis(3)
K = ng.make_axis(11)
axes_a = ng.make_axes([H, W, C])
a = ng.constant(np.ones(axes_a.lengths), axes=axes_a)
flat_a = ng.flatten_at(a, 2)
axes_b = ng.make_axes([C, K])
b = ng.constant(np.ones(axes_b.lengths), axes=axes_b)
result = ng.dot(b, flat_a)
with ExecutorFactory() as factory:
result_fun = factory.executor(result)
result_val = result_fun()
result_correct = np.ones_like(result_val) * C.length
ng.testing.assert_allclose(result_val, result_correct)
def __call__(self, input_data, weights):
'''
input_data in this case is a numpy array with batch_size on axis 1
and weights is a matrix with 1 column
'''
if self.state is None:
self.state = np.ones_like(weights)
if self.velocity is None:
self.velocity = np.zeros_like(weights)
gradient = - input_data.mean(axis=1)
self.state[:] = self.decay_rate * self.state + \
(1.0 - self.decay_rate) * np.square(gradient)
self.velocity = self.velocity * self.momentum + \
self.learning_rate * gradient / np.sqrt(self.state + self.epsilon) + \
self.learning_rate * self.wdecay * weights
weights[:] = weights - self.velocity
return weights
def Affine_test(self,N,sizex,sizey,sizez,times,stop_time,typeofT,colors):
for i in range(times):
# Theta
idx = np.random.uniform(-1, 1);idy = np.random.uniform(-1, 1);idz = np.random.uniform(-1, 1)
swithx = np.random.uniform(0,1);swithy = np.random.uniform(0,1);swithz = np.random.uniform(0,1)
rotatex = np.random.uniform(-1, 1);rotatey = np.random.uniform(-1, 1);rotatez = np.random.uniform(-1, 1)
cx = np.array([idx,rotatey,rotatez,swithx]);cy = np.array([rotatex,idy,rotatez,swithy]);cz = np.array([rotatex,rotatey,idz,swithz])
# Source Grid
x = np.linspace(-sizex, sizex, N);y = np.linspace(-sizey, sizey, N);z = np.linspace(-sizez, sizez, N)
x, y, z = np.meshgrid(x, y, z)
xgs, ygs, zgs = x.flatten(), y.flatten(),z.flatten()
gps = np.vstack([xgs, ygs, zgs, np.ones_like(xgs)]).T
# transform
xgt = np.dot(gps, cx);ygt = np.dot(gps, cy);zgt = np.dot(gps, cz)
# display
showIm = ShowImage()
showIm.Show_transform(xgs,ygs,zgs,xgt,ygt,zgt,sizex,sizey,sizez,stop_time,typeofT,N,colors)
def pseudo_call_gene_low_level(simulationTime=None, rho=None, temperatureHat=None, densityHat=None, safetyFactor=None,
ionMass=1, ionCharge=1, Lref=None, Bref=None, rhoStar=None, Tref=None, nref=None, checkpointSuffix=0):
"""Function to emulate a call to GENE with the same input arguments and return values.
Used for testing other code when the overhead of an actual startup of GENE is not needed. Of course, this can only
be used to test correctness of syntax, not correctness of values.
"""
# check inputs have been provided
for var in (simulationTime, rho, temperatureHat, densityHat, safetyFactor, Lref, Bref, rhoStar, Tref, nref):
if var is None:
#logging.error("Input variables must be provided in call_gene_low_level.")
raise ValueError
MPIrank = 1
dVdxHat = np.ones_like(rho)
sqrt_gxx = np.ones_like(rho)
avgParticleFluxHat = np.ones_like(rho)
avgHeatFluxHat = np.ones_like(rho)
temperatureOutput = np.ones_like(rho)
densityOutput = np.ones_like(rho)
return (MPIrank, dVdxHat, sqrt_gxx, avgParticleFluxHat, avgHeatFluxHat, temperatureOutput, densityOutput)
def _add_noise(v, amplitude):
"""Add noise to an array v in the following way:.
noisy_v = (1+h) * v
where h is a random noise with specified standard deviation. The noise h is trimmed to
be zero close to both boundaries.
Inputs:
v input to add noise to (array)
ampltitude specified standard deviation of noise (scalar)
tac autocorrelation time measured in discrete samples (scalar)
Outputs:
noisy_v v with noise
"""
# generate noise that is constant throughout space
h = np.random.normal(scale=amplitude) * np.ones_like(v)
# damped the sides of the noise close to the boundaries
h = dampen_sides(h)
noisy_v = (1 + h) * v
return noisy_v
def __call__(self, t, x, n):
# Define the contributions to the H coefficients for the Shestakov Problem
H1 = np.ones_like(x)
#H7 = shestakov_nonlinear_diffusion.H7contrib_Source(x)
H7 = source(x)
(H2turb, H3, extradata) = self.turbhandler.Hcontrib_turbulent_flux(n)
H4 = None
H6 = None
# add "other" diffusive contributions by specifying a diffusivity, H2 = V'D [but V' = 1 here]
H2constdiff = 0.03
def diffusivity_right(x):
diffusivity = np.zeros_like(x)
xr = 0.85
D0 = 7
diffusivity[x > xr] = D0
return diffusivity
H2 = H2turb + H2constdiff
#H2 = H2turb + H2constdiff + diffusivity_right(x) # if adding const to right edge
return (H1, H2, H3, H4, H6, H7, extradata)
def __call__(self, t, x, n):
# Define the contributions to the H coefficients for the Shestakov Problem
H1 = np.ones_like(x)
#H7 = shestakov_nonlinear_diffusion.H7contrib_Source(x)
H7 = source(x)
(H2turb, H3, extradata) = self.turbhandler.Hcontrib_turbulent_flux(n)
H4 = None
H6 = None
# add "other" diffusive contributions by specifying a diffusivity, H2 = V'D [but V' = 1 here]
H2constdiff = 0.03
def diffusivity_right(x):
diffusivity = np.zeros_like(x)
xr = 0.85
D0 = 7
diffusivity[x > xr] = D0
return diffusivity
H2 = H2turb + H2constdiff
#H2 = H2turb + H2constdiff + diffusivity_right(x) # if adding const to right edge
return (H1, H2, H3, H4, H6, H7, extradata)
def setup_parameters_different_grids_tango_inside():
# set up radial grids with Tango's outer radial boundary radially inward that of GENE.
simulationTime = 0.4
Lref = 1.65
Bref = 2.5
majorRadius = 1.65
minorRadius = 0.594
rhoStar = 1/140
checkpointSuffix = 999
numRadialPtsTango = 100
numRadialPtsGene = 80
rhoTango = np.linspace(0.1, 0.8, numRadialPtsTango) # rho = r/a
rhoGene = np.linspace(0.2, 0.9, numRadialPtsGene)
rTango = rhoTango * minorRadius # physical radius r
rGene = rhoGene * minorRadius
safetyFactorGeneGrid = tango.parameters.analytic_safety_factor(rGene, minorRadius, majorRadius)
e = 1.60217662e-19 # electron charge
temperatureGeneGrid = 1000 * e * np.ones_like(rGene)
densityTangoGrid = 1e19 * np.ones_like(rTango)
densityGeneGrid = 1e19 * np.ones_like(rGene)
gridMapper = tango.interfacegrids_gene.GridInterfaceTangoInside(rTango, rGene)
return (simulationTime, rTango, rGene, temperatureGeneGrid, densityTangoGrid, densityGeneGrid, safetyFactorGeneGrid, Lref, Bref, majorRadius, minorRadius, rhoStar, gridMapper, checkpointSuffix)
def setup_parameters_different_grids_tango_outside():
# set up radial grids with Tango's outer radial boundary radially inward that of GENE.
simulationTime = 0.4
Lref = 1.65
Bref = 2.5
majorRadius = 1.65
minorRadius = 0.594
rhoStar = 1/140
checkpointSuffix = 999
numRadialPtsTango = 100
numRadialPtsGene = 80
rhoTango = np.linspace(0.1, 0.9, numRadialPtsTango) # rho = r/a
rhoGene = np.linspace(0.2, 0.7, numRadialPtsGene)
rTango = rhoTango * minorRadius # physical radius r
rGene = rhoGene * minorRadius
safetyFactorGeneGrid = tango.parameters.analytic_safety_factor(rGene, minorRadius, majorRadius)
e = 1.60217662e-19 # electron charge
temperatureGeneGrid = 1000 * e * np.ones_like(rGene)
densityTangoGrid = 1e19 * np.ones_like(rTango)
densityGeneGrid = 1e19 * np.ones_like(rGene)
gridMapper = tango.interfacegrids_gene.GridInterfaceTangoOutside(rTango, rGene)
return (simulationTime, rTango, rGene, temperatureGeneGrid, densityTangoGrid, densityGeneGrid, safetyFactorGeneGrid, Lref, Bref, majorRadius, minorRadius, rhoStar, gridMapper, checkpointSuffix)
def test_energy_conservation_sech2disk_manyparticles():
# Test that energy is conserved for a self-gravitating disk
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
g= wendy.nbody(x,v,m,0.05)
E= wendy.energy(x,v,m)
cnt= 0
while cnt < 100:
tx,tv= next(g)
assert numpy.fabs(wendy.energy(tx,tv,m)-E) < 10.**-10., "Energy not conserved during simple N-body integration"
cnt+= 1
return None
def test_energy_conservation_sech2disk_manyparticles():
# Test that energy is conserved for a self-gravitating disk
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
omega= 1.1
g= wendy.nbody(x,v,m,0.05,omega=omega)
E= wendy.energy(x,v,m,omega=omega)
cnt= 0
while cnt < 100:
tx,tv= next(g)
assert numpy.fabs(wendy.energy(tx,tv,m,omega=omega)-E) < 10.**-10., "Energy not conserved during simple N-body integration with external harmonic potential"
cnt+= 1
return None
def test_energy_conservation_sech2disk_manyparticles():
# Test that energy is conserved for a self-gravitating disk
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
g= wendy.nbody(x,v,m,0.05,approx=True,nleap=1000)
E= wendy.energy(x,v,m)
cnt= 0
while cnt < 100:
tx,tv= next(g)
assert numpy.fabs(wendy.energy(tx,tv,m)-E)/E < 10.**-6., "Energy not conserved during approximate N-body integration"
cnt+= 1
return None
def test_notracermasses():
# approx should work with tracer sheets
# Test that energy is conserved for a self-gravitating disk
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
m[N//2:]= 0.
m*= 2.
g= wendy.nbody(x,v,m,0.05,approx=True,nleap=1000)
E= wendy.energy(x,v,m)
cnt= 0
while cnt < 100:
tx,tv= next(g)
assert numpy.fabs(wendy.energy(tx,tv,m)-E)/E < 10.**-6., "Energy not conserved during approximate N-body integration with some tracer particles"
cnt+= 1
return None
def test_energy_conservation_sech2disk_manyparticles():
# Test that energy is conserved for a self-gravitating disk
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
omega= 1.1
g= wendy.nbody(x,v,m,0.05,omega=omega,approx=True,nleap=1000)
E= wendy.energy(x,v,m,omega=omega)
cnt= 0
while cnt < 100:
tx,tv= next(g)
assert numpy.fabs(wendy.energy(tx,tv,m,omega=omega)-E)/E < 10.**-6., "Energy not conserved during approximate N-body integration with external harmonic potential"
cnt+= 1
return None
def test_againstexact_sech2disk_manyparticles():
# Test that the exact N-body and the approximate N-body agree
N= 101
totmass= 1.
sigma= 1.
zh= 2.*sigma**2./totmass
x= numpy.arctanh(2.*numpy.random.uniform(size=N)-1)*zh
v= numpy.random.normal(size=N)*sigma
v-= numpy.mean(v) # stabilize
m= numpy.ones_like(x)/N*(1.+0.1*(2.*numpy.random.uniform(size=N)-1))
omega= 1.1
g= wendy.nbody(x,v,m,0.05,approx=True,nleap=2000,omega=omega)
ge= wendy.nbody(x,v,m,0.05,omega=omega)
cnt= 0
while cnt < 100:
tx,tv= next(g)
txe,tve= next(ge)
assert numpy.all(numpy.fabs(tx-txe) < 10.**-5.), "Exact and approximate N-body give different positions"
assert numpy.all(numpy.fabs(tv-tve) < 10.**-5.), "Exact and approximate N-body give different positions"
cnt+= 1
return None
def gain_substitution_scalar(gain, x, xwt):
nants, nchan, nrec, _ = gain.shape
newgain = numpy.ones_like(gain, dtype='complex')
gwt = numpy.zeros_like(gain, dtype='float')
# We are going to work with Jones 2x2 matrix formalism so everything has to be
# converted to that format
x = x.reshape(nants, nants, nchan, nrec, nrec)
xwt = xwt.reshape(nants, nants, nchan, nrec, nrec)
for ant1 in range(nants):
for chan in range(nchan):
# Loop over e.g. 'RR', 'LL, or 'xx', 'YY' ignoring cross terms
top = numpy.sum(x[:, ant1, chan, 0, 0] *
gain[:, chan, 0, 0] * xwt[:, ant1, chan, 0, 0], axis=0)
bot = numpy.sum((gain[:, chan, 0, 0] * numpy.conjugate(gain[:, chan, 0, 0]) *
xwt[:, ant1, chan, 0, 0]).real, axis=0)
if bot > 0.0:
newgain[ant1, chan, 0, 0] = top / bot
gwt[ant1, chan, 0, 0] = bot
else:
newgain[ant1, chan, 0, 0] = 0.0
gwt[ant1, chan, 0, 0] = 0.0
return newgain, gwt
def test_balance_weights(self):
labels = [1, 1, -1, -1, -1]
predictions = [-1, -1, 1, 1, 1] # all errors
default_weights = np.abs(center(np.asarray(labels)))
exp_error = balanced_classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions, default_weights)
assert_equals(exp_error, error)
null_weights = np.ones_like(labels)
exp_error = classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions, null_weights)
assert_equals(exp_error, error)
# Balanced classes
labels = [1, 1, 1, -1, -1, -1]
predictions = [-1, -1, -1, 1, 1, 1] # all errors
exp_error = classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions)
assert_equals(exp_error, error)
def test_balance_weights():
"""Test balanced classification error with custom weights."""
labels = [1, 1, -1, -1, -1]
predictions = [-1, -1, 1, 1, 1] # all errors
default_weights = np.abs(center(np.asarray(labels)))
exp_error = balanced_classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions, default_weights)
assert_equals(exp_error, error)
null_weights = np.ones_like(labels)
exp_error = classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions, null_weights)
assert_equals(exp_error, error)
# Balanced classes
labels = [1, 1, 1, -1, -1, -1]
predictions = [-1, -1, -1, 1, 1, 1] # all errors
exp_error = classification_error(labels, predictions)
error = balanced_classification_error(labels, predictions)
assert_equals(exp_error, error)
def initialize_labels(self, Y):
y_nodes_flat = [y_val for y in Y for y_val in y.nodes]
y_links_flat = [y_val for y in Y for y_val in y.links]
self.prop_encoder_ = LabelEncoder().fit(y_nodes_flat)
self.link_encoder_ = LabelEncoder().fit(y_links_flat)
self.n_prop_states = len(self.prop_encoder_.classes_)
self.n_link_states = len(self.link_encoder_.classes_)
self.prop_cw_ = np.ones_like(self.prop_encoder_.classes_,
dtype=np.double)
self.link_cw_ = compute_class_weight(self.class_weight,
self.link_encoder_.classes_,
y_links_flat)
self.link_cw_ /= self.link_cw_.min()
logging.info('Setting node class weights {}'.format(", ".join(
"{}: {}".format(lbl, cw) for lbl, cw in zip(
self.prop_encoder_.classes_, self.prop_cw_))))
logging.info('Setting link class weights {}'.format(", ".join(
"{}: {}".format(lbl, cw) for lbl, cw in zip(
self.link_encoder_.classes_, self.link_cw_))))
def get_adagrad(learning_rate=0.5):
"""
Adaptive Subgradient Methods for Online Learning and Stochastic Optimization
John Duchi, Elad Hazan and Yoram Singer, Journal of Machine Learning Research 12 (2011) 2121-2159
http://www.jmlr.org/papers/volume12/duchi11a/duchi11a.pdf
"""
sum_square_gradient = None
def adagrad(gradient):
nonlocal sum_square_gradient
if sum_square_gradient is None:
sum_square_gradient = np.ones_like(gradient)
sum_square_gradient += gradient ** 2
return learning_rate / np.sqrt(sum_square_gradient)
return adagrad
def load_data(self):
# Create the data using magic numbers to approximate the figure in
# canevet_icml2016
x = np.linspace(0, 1, self.N).astype(np.float32)
ones = np.ones_like(x).astype(int)
boundary = np.sin(4*(x + 0.5)**5)/3 + 0.5
data = np.empty(shape=[self.N, self.N, 3], dtype=np.float32)
data[:, :, 0] = 1-x
for i in range(self.N):
data[i, :, 1] = 1-x[i]
data[i, :, 2] = 1 / (1 + np.exp(self.smooth*(x - boundary[i])))
data[i, :, 2] = np.random.binomial(ones, data[i, :, 2])
data = data.reshape(-1, 3)
np.random.shuffle(data)
# Create train and test arrays
split = int(len(data)*self.test_split)
X_train = data[:-split, :2]
y_train = data[:-split, 2]
X_test = data[-split:, :2]
y_test = data[-split:, 2]
return (X_train, y_train), (X_test, y_test)
def __init__(self, dataset, reweighting, model, large_batch=1024,
forward_batch_size=128, steps_per_epoch=300, recompute=2,
s_e=(1, 1), n_epochs=1):
super(OnlineBatchSelectionSampler, self).__init__(
dataset,
reweighting,
model,
large_batch=large_batch,
forward_batch_size=forward_batch_size
)
# The configuration of OnlineBatchSelection
self.steps_per_epoch = steps_per_epoch
self.recompute = recompute
self.s_e = s_e
self.n_epochs = n_epochs
# Mutable variables to be updated
self._batch = 0
self._epoch = 0
self._raw_scores = np.ones((len(dataset.train_data),))
self._scores = np.ones_like(self._raw_scores)
self._ranks = np.arange(len(dataset.train_data))
def load_data(self):
# Create the data using magic numbers to approximate the figure in
# canevet_icml2016
x = np.linspace(0, 1, self.N).astype(np.float32)
ones = np.ones_like(x).astype(int)
boundary = np.sin(4*(x + 0.5)**5)/3 + 0.5
data = np.empty(shape=[self.N, self.N, 3], dtype=np.float32)
data[:, :, 0] = 1-x
for i in range(self.N):
data[i, :, 1] = 1-x[i]
data[i, :, 2] = 1 / (1 + np.exp(self.smooth*(x - boundary[i])))
data[i, :, 2] = np.random.binomial(ones, data[i, :, 2])
data = data.reshape(-1, 3)
np.random.shuffle(data)
# Create train and test arrays
split = int(len(data)*self.test_split)
X_train = data[:-split, :2]
y_train = data[:-split, 2]
X_test = data[-split:, :2]
y_test = data[-split:, 2]
return (X_train, y_train), (X_test, y_test)
def __init__(self, dataset, reweighting, model, large_batch=1024,
forward_batch_size=128, steps_per_epoch=300, recompute=2,
s_e=(1, 1), n_epochs=1):
super(OnlineBatchSelectionSampler, self).__init__(
dataset,
reweighting,
model,
large_batch=large_batch,
forward_batch_size=forward_batch_size
)
# The configuration of OnlineBatchSelection
self.steps_per_epoch = steps_per_epoch
self.recompute = recompute
self.s_e = s_e
self.n_epochs = n_epochs
# Mutable variables to be updated
self._batch = 0
self._epoch = 0
self._raw_scores = np.ones((len(dataset.train_data),))
self._scores = np.ones_like(self._raw_scores)
self._ranks = np.arange(len(dataset.train_data))
def load_data(self):
# Create the data using magic numbers to approximate the figure in
# canevet_icml2016
x = np.linspace(0, 1, self.N).astype(np.float32)
ones = np.ones_like(x).astype(int)
boundary = np.sin(4*(x + 0.5)**5)/3 + 0.5
data = np.empty(shape=[self.N, self.N, 3], dtype=np.float32)
data[:, :, 0] = 1-x
for i in range(self.N):
data[i, :, 1] = 1-x[i]
data[i, :, 2] = 1 / (1 + np.exp(self.smooth*(x - boundary[i])))
data[i, :, 2] = np.random.binomial(ones, data[i, :, 2])
data = data.reshape(-1, 3)
np.random.shuffle(data)
# Create train and test arrays
split = int(len(data)*self.test_split)
X_train = data[:-split, :2]
y_train = data[:-split, 2]
X_test = data[-split:, :2]
y_test = data[-split:, 2]
return (X_train, y_train), (X_test, y_test)
def load_data(self):
# Create the data using magic numbers to approximate the figure in
# canevet_icml2016
x = np.linspace(0, 1, self.N).astype(np.float32)
ones = np.ones_like(x).astype(int)
boundary = np.sin(4*(x + 0.5)**5)/3 + 0.5
data = np.empty(shape=[self.N, self.N, 3], dtype=np.float32)
data[:, :, 0] = 1-x
for i in range(self.N):
data[i, :, 1] = 1-x[i]
data[i, :, 2] = 1 / (1 + np.exp(self.smooth*(x - boundary[i])))
data[i, :, 2] = np.random.binomial(ones, data[i, :, 2])
data = data.reshape(-1, 3)
np.random.shuffle(data)
# Create train and test arrays
split = int(len(data)*self.test_split)
X_train = data[:-split, :2]
y_train = data[:-split, 2]
X_test = data[-split:, :2]
y_test = data[-split:, 2]
return (X_train, y_train), (X_test, y_test)
def __init__(self, dataset, reweighting, model, large_batch=1024,
forward_batch_size=128, steps_per_epoch=300, recompute=2,
s_e=(1, 1), n_epochs=1):
super(OnlineBatchSelectionSampler, self).__init__(
dataset,
reweighting,
model,
large_batch=large_batch,
forward_batch_size=forward_batch_size
)
# The configuration of OnlineBatchSelection
self.steps_per_epoch = steps_per_epoch
self.recompute = recompute
self.s_e = s_e
self.n_epochs = n_epochs
# Mutable variables to be updated
self._batch = 0
self._epoch = 0
self._raw_scores = np.ones((len(dataset.train_data),))
self._scores = np.ones_like(self._raw_scores)
self._ranks = np.arange(len(dataset.train_data))
def get_data(discrete_time):
y_test, y_train, u_train = generate_weibull(A=real_a,
B=real_b,
# <np.inf -> impose censoring
C=censoring_point,
shape=[n_sequences,
n_timesteps, 1],
discrete_time=discrete_time)
# With random input it _should_ learn weight 0
x_train = x_test = np.random.uniform(
low=-1, high=1, size=[n_sequences, n_timesteps, n_features])
# y_test is uncencored data
y_test = np.append(y_test, np.ones_like(y_test), axis=-1)
y_train = np.append(y_train, u_train, axis=-1)
return y_train, x_train, y_test, x_test
def test_ignore_nans(self):
""" Test that NaNs are ignored. """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
source.append(np.full_like(source[0], np.nan))
product = cprod(source, ignore_nan = True)
self.assertTrue(np.allclose(product, np.ones_like(product)))
def test_dtype(self):
""" Test that dtype argument is working """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
product = cprod(source, dtype = np.int)
self.assertTrue(np.allclose(product, np.ones_like(product)))
self.assertEqual(product.dtype, np.int)
def test_trivial(self):
""" Test a product of ones """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
product = last(iprod(source))
self.assertTrue(np.allclose(product, np.ones_like(product)))
def test_ignore_nans(self):
""" Test that NaNs are ignored. """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
source.append(np.full_like(source[0], np.nan))
product = last(iprod(source, ignore_nan = True))
self.assertTrue(np.allclose(product, np.ones_like(product)))
def test_dtype(self):
""" Test that dtype argument is working """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
product = last(iprod(source, dtype = np.int))
self.assertTrue(np.allclose(product, np.ones_like(product)))
self.assertEqual(product.dtype, np.int)
def test_trivial(self):
""" Test a product of ones """
source = [np.ones((16,), dtype = np.float) for _ in range(10)]
product = last(inanprod(source))
self.assertTrue(np.allclose(product, np.ones_like(product)))
def composed_triangular_mesh(triangular_mesh_dict):
start_time = time()
print "--> Composing triangular mesh..."
mesh = TriangularMesh()
triangle_cell_matching = {}
mesh_points = np.concatenate([triangular_mesh_dict[c].points.keys() for c in triangular_mesh_dict.keys()])
mesh_point_positions = np.concatenate([triangular_mesh_dict[c].points.values() for c in triangular_mesh_dict.keys()])
mesh.points = dict(zip(mesh_points,mesh_point_positions))
mesh_triangles = np.concatenate([triangular_mesh_dict[c].triangles.values() for c in triangular_mesh_dict.keys()])
mesh.triangles = dict(zip(np.arange(len(mesh_triangles)),mesh_triangles))
mesh_cells = np.concatenate([c*np.ones_like(triangular_mesh_dict[c].triangles.keys()) for c in triangular_mesh_dict.keys()])
triangle_cell_matching = dict(zip(np.arange(len(mesh_triangles)),mesh_cells))
# for c in triangular_mesh_dict.keys():
# cell_start_time = time()
# cell_mesh = triangular_mesh_dict[c]
# # mesh_point_max_id = np.max(mesh.points.keys()) if len(mesh.points)>0 else 0
# mesh.points.update(cell_mesh.points)
# if len(cell_mesh.triangles)>0:
# mesh_triangle_max_id = np.max(mesh.triangles.keys()) if len(mesh.triangles)>0 else 0
# mesh.triangles.update(dict(zip(list(np.array(cell_mesh.triangles.keys())+mesh_triangle_max_id),cell_mesh.triangles.values())))
# triangle_cell_matching.update(dict(zip(list(np.array(cell_mesh.triangles.keys())+mesh_triangle_max_id),[c for f in cell_mesh.triangles])))
# cell_end_time = time()
# print " --> Adding cell ",c," (",len(cell_mesh.triangles)," triangles ) [",cell_end_time-cell_start_time,"s]"
end_time = time()
print "<-- Composing triangular mesh [",end_time-start_time,"]"
return mesh, triangle_cell_matching
def _loess_predict(X, y_tr, X_pred, bandwidth):
X_tr = np.column_stack((np.ones_like(X), X))
X_te = np.column_stack((np.ones_like(X_pred), X_pred))
y_te = []
for x in X_te:
ws = np.exp(-np.sum((X_tr - x)**2, axis=1) / (2 * bandwidth**2))
W = scipy.sparse.dia_matrix((ws, 0), shape=(X_tr.shape[0],) * 2)
theta = np.linalg.pinv(X_tr.T.dot(W.dot(X_tr))).dot(X_tr.T.dot(W.dot(y_tr)))
y_te.append(np.dot(x, theta))
return np.array(y_te)
def __init__(self, X, Y, batch_size, cropsize=0):
assert len(X) == len(Y), 'X and Y must be the same length {}!={}'.format(len(X),len(Y))
print('starting balanced generator')
self.X = X
self.Y = Y
self.cropsize=cropsize
self.batch_size = int(batch_size)
self.pmatrix = np.ones_like(self.Y)
self.reset()
def gamma_fullsum_grad(
gamma,
node_vec,
eventmemes,
etimes,
T,
mu,
alpha,
omega,
W,
beta,
kernel_evaluate,
K_evaluate,
):
'''
it actually returns negated gradient.
'''
gradres = np.ones_like(gamma) * -T * np.sum(mu)
for (eventidx, (etime1, infected_u, eventmeme)) in \
enumerate(izip(etimes, node_vec, eventmemes)):
gradres[eventmeme] += mu[infected_u] \
/ np.exp(event_nonapproximated_logintensity(
infected_u,
eventmeme,
etime1,
T,
etimes[:eventidx],
node_vec[:eventidx],
eventmemes[:eventidx],
mu,
gamma,
omega,
alpha,
kernel_evaluate,
))
return -gradres
# =====