Commit 6a8e4c3b authored by Reinhard Prix's avatar Reinhard Prix

converted to python3 via '2to3 -w'

- tests currently fail
parent aa700d85
......@@ -28,13 +28,13 @@ data.make_data()
# The predicted twoF, given by lalapps_predictFstat can be accessed by
twoF = data.predict_fstat()
print 'Predicted twoF value: {}\n'.format(twoF)
print('Predicted twoF value: {}\n'.format(twoF))
DeltaF0 = 1e-7
DeltaF1 = 1e-13
VF0 = (np.pi * duration * DeltaF0)**2 / 3.0
VF1 = (np.pi * duration**2 * DeltaF1)**2 * 4/45.
print '\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1)
print('\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1))
theta_prior = {'F0': {'type': 'unif',
'lower': F0-DeltaF0/2.,
......
......@@ -28,13 +28,13 @@ data.make_data()
# The predicted twoF, given by lalapps_predictFstat can be accessed by
twoF = data.predict_fstat()
print 'Predicted twoF value: {}\n'.format(twoF)
print('Predicted twoF value: {}\n'.format(twoF))
DeltaF0 = 1e-7
DeltaF1 = 1e-13
VF0 = (np.pi * duration * DeltaF0)**2 / 3.0
VF1 = (np.pi * duration**2 * DeltaF1)**2 * 4/45.
print '\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1)
print('\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1))
theta_prior = {'F0': {'type': 'unif',
'lower': F0-DeltaF0/2.,
......
......@@ -28,7 +28,7 @@ data.make_data()
# The predicted twoF, given by lalapps_predictFstat can be accessed by
twoF = data.predict_fstat()
print 'Predicted twoF value: {}\n'.format(twoF)
print('Predicted twoF value: {}\n'.format(twoF))
# Search
VF0 = VF1 = 1e5
......
......@@ -66,5 +66,5 @@ mcmc.plot_corner(label_offset=0.25, truths=[0, 0, 0, 0],
mcmc.print_summary()
print('Prior widths =', F0_width, F1_width)
print("Actual run time = {}".format(dT))
print(('Prior widths =', F0_width, F1_width))
print(("Actual run time = {}".format(dT)))
......@@ -62,6 +62,6 @@ fig.savefig('{}/{}_projection_matrix.png'.format(outdir, label),
bbox_inches='tight')
print('Prior widths =', F0_width, F1_width)
print("Actual run time = {}".format(dT))
print("Actual number of grid points = {}".format(search.data.shape[0]))
print(('Prior widths =', F0_width, F1_width))
print(("Actual run time = {}".format(dT)))
print(("Actual number of grid points = {}".format(search.data.shape[0])))
......@@ -27,13 +27,13 @@ data.make_data()
# The predicted twoF, given by lalapps_predictFstat can be accessed by
twoF = data.predict_fstat()
print 'Predicted twoF value: {}\n'.format(twoF)
print('Predicted twoF value: {}\n'.format(twoF))
DeltaF0 = 1e-7
DeltaF1 = 1e-13
VF0 = (np.pi * duration * DeltaF0)**2 / 3.0
VF1 = (np.pi * duration**2 * DeltaF1)**2 * 4/45.
print '\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1)
print('\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1))
theta_prior = {'F0': {'type': 'unif',
'lower': F0-DeltaF0/2.,
......
......@@ -28,13 +28,13 @@ data.make_data()
# The predicted twoF, given by lalapps_predictFstat can be accessed by
twoF = data.predict_fstat()
print 'Predicted twoF value: {}\n'.format(twoF)
print('Predicted twoF value: {}\n'.format(twoF))
DeltaF0 = 1e-7
DeltaF1 = 1e-13
VF0 = (np.pi * duration * DeltaF0)**2 / 3.0
VF1 = (np.pi * duration**2 * DeltaF1)**2 * 4/45.
print '\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1)
print('\nV={:1.2e}, VF0={:1.2e}, VF1={:1.2e}\n'.format(VF0*VF1, VF0, VF1))
theta_prior = {'F0': {'type': 'unif',
'lower': F0-DeltaF0/2.,
......
from __future__ import division as _division
from .core import BaseSearchClass, ComputeFstat, SemiCoherentSearch, SemiCoherentGlitchSearch
from .make_sfts import Writer, GlitchWriter, FrequencyModulatedArtifactWriter, FrequencyAmplitudeModulatedArtifactWriter
......
""" The core tools used in pyfstat """
from __future__ import division, absolute_import, print_function
import os
import logging
......
""" Searches using grid-based methods """
from __future__ import division, absolute_import, print_function
import os
import logging
......@@ -342,7 +342,7 @@ class GridSearch(BaseSearchClass):
def print_max_twoF(self):
d = self.get_max_twoF()
print('Max twoF values for {}:'.format(self.label))
for k, v in d.iteritems():
for k, v in d.items():
print(' {}={}'.format(k, v))
def set_out_file(self, extra_label=None):
......@@ -1006,7 +1006,7 @@ class EarthTest(GridSearch):
vals = [self.minStartTime, self.maxStartTime, self.F0, self.F1,
self.F2, self.Alpha, self.Delta]
self.special_data = {'zero': [0, 0, 0]}
for key, (dR, dphi, dP) in self.special_data.iteritems():
for key, (dR, dphi, dP) in self.special_data.items():
rescaleRadius = (1 + dR / lal.REARTH_SI)
rescalePeriod = (1 + dP / lal.DAYSID_SI)
lalpulsar.BarycenterModifyEarthRotation(
......
......@@ -110,7 +110,7 @@ def get_ephemeris_files():
logging.warning('No [earth/sun]_ephem found in '+config_file+'. '+please)
earth_ephem = None
sun_ephem = None
elif env_var in os.environ.keys():
elif env_var in list(os.environ.keys()):
earth_ephem = os.path.join(os.environ[env_var],'earth00-40-DE421.dat.gz')
sun_ephem = os.path.join(os.environ[env_var],'sun00-40-DE421.dat.gz')
if not ( os.path.isfile(earth_ephem) and os.path.isfile(sun_ephem) ):
......
""" pyfstat tools to generate sfts """
from __future__ import division, absolute_import, print_function
import numpy as np
import logging
......@@ -477,7 +477,7 @@ class FrequencyModulatedArtifactWriter(Writer):
linePhi = 0
lineFreq_old = 0
for i in tqdm(range(self.nsfts)):
for i in tqdm(list(range(self.nsfts))):
mid_time = self.tstart + (i+.5)*self.Tsft
lineFreq = self.get_frequency(mid_time)
......@@ -517,7 +517,7 @@ class FrequencyModulatedArtifactWriter(Writer):
logging.info('Using {} threads'.format(args.N))
try:
with pathos.pools.ProcessPool(args.N) as p:
list(tqdm(p.imap(self.make_ith_sft, range(self.nsfts)),
list(tqdm(p.imap(self.make_ith_sft, list(range(self.nsfts))),
total=self.nsfts))
except KeyboardInterrupt:
p.terminate()
......@@ -525,7 +525,7 @@ class FrequencyModulatedArtifactWriter(Writer):
logging.info(
"No multiprocessing requested or `pathos` not install, cont."
" without multiprocessing")
for i in tqdm(range(self.nsfts)):
for i in tqdm(list(range(self.nsfts))):
self.make_ith_sft(i)
self.concatenate_sft_files()
......
""" Searches using MCMC-based methods """
from __future__ import division, absolute_import, print_function
import sys
import os
......@@ -221,7 +221,7 @@ class MCMCSearch(core.BaseSearchClass):
self.theta_keys = []
fixed_theta_dict = {}
for key, val in self.theta_prior.iteritems():
for key, val in self.theta_prior.items():
if type(val) is dict:
fixed_theta_dict[key] = 0
self.theta_keys.append(key)
......@@ -953,7 +953,7 @@ class MCMCSearch(core.BaseSearchClass):
See the pyfstat.core.plot_twoF_cumulative function for further details
"""
d, maxtwoF = self.get_max_twoF()
for key, val in self.theta_prior.iteritems():
for key, val in self.theta_prior.items():
if key not in d:
d[key] = val
......@@ -1223,8 +1223,8 @@ class MCMCSearch(core.BaseSearchClass):
def _generate_scattered_p0(self, p):
""" Generate a set of p0s scattered about p """
p0 = [[p + self.scatter_val * p * np.random.randn(self.ndim)
for i in xrange(self.nwalkers)]
for j in xrange(self.ntemps)]
for i in range(self.nwalkers)]
for j in range(self.ntemps)]
return p0
def _generate_initial_p0(self):
......@@ -1349,7 +1349,7 @@ class MCMCSearch(core.BaseSearchClass):
setattr(self, key, new_d[key])
mod_keys = []
for key in new_d.keys():
for key in list(new_d.keys()):
if key in old_d:
if new_d[key] != old_d[key]:
mod_keys.append((key, old_d[key], new_d[key]))
......@@ -1486,10 +1486,10 @@ class MCMCSearch(core.BaseSearchClass):
if hasattr(self, 'theta0_index'):
f.write('theta0_index = {}\n'.format(self.theta0_idx))
if method == 'med':
for key, val in median_std_d.iteritems():
for key, val in median_std_d.items():
f.write('{} = {:1.16e}\n'.format(key, val))
if method == 'twoFmax':
for key, val in max_twoF_d.iteritems():
for key, val in max_twoF_d.items():
f.write('{} = {:1.16e}\n'.format(key, val))
def generate_loudest(self):
......@@ -1514,7 +1514,7 @@ class MCMCSearch(core.BaseSearchClass):
f.write(r"\begin{tabular}{c l c} \hline" + '\n'
r"Parameter & & & \\ \hhline{====}")
for key, prior in self.theta_prior.iteritems():
for key, prior in self.theta_prior.items():
if type(prior) is dict:
Type = prior['type']
if Type == "unif":
......@@ -1546,10 +1546,10 @@ class MCMCSearch(core.BaseSearchClass):
if hasattr(self, 'theta0_idx'):
logging.info('theta0 index: {}'.format(self.theta0_idx))
logging.info('Max twoF: {} with parameters:'.format(max_twoF))
for k in np.sort(max_twoFd.keys()):
for k in np.sort(list(max_twoFd.keys())):
print(' {:10s} = {:1.9e}'.format(k, max_twoFd[k]))
logging.info('Median +/- std for production values')
for k in np.sort(median_std_d.keys()):
for k in np.sort(list(median_std_d.keys())):
if 'std' not in k:
logging.info(' {:10s} = {:1.9e} +/- {:1.9e}'.format(
k, median_std_d[k], median_std_d[k+'_std']))
......@@ -1668,7 +1668,7 @@ class MCMCSearch(core.BaseSearchClass):
def write_evidence_file_from_dict(self, EvidenceDict, evidence_file_name):
with open(evidence_file_name, 'w+') as f:
for key, val in EvidenceDict.iteritems():
for key, val in EvidenceDict.items():
f.write('{} {} {}\n'.format(key, val[0], val[1]))
......@@ -1801,7 +1801,7 @@ class MCMCGlitchSearch(MCMCSearch):
r'$\delta$'] + full_glitch_symbols)
self.theta_keys = []
fixed_theta_dict = {}
for key, val in self.theta_prior.iteritems():
for key, val in self.theta_prior.items():
if type(val) is dict:
fixed_theta_dict[key] = 0
if key in glitch_keys:
......@@ -1863,7 +1863,7 @@ class MCMCGlitchSearch(MCMCSearch):
fig, ax = plt.subplots()
d, maxtwoF = self.get_max_twoF()
for key, val in self.theta_prior.iteritems():
for key, val in self.theta_prior.items():
if key not in d:
d[key] = val
......@@ -2223,7 +2223,7 @@ class MCMCFollowUpSearch(MCMCSemiCoherentSearch):
def check_old_run_setup(self, old_setup, **kwargs):
try:
truths = [val == old_setup[key] for key, val in kwargs.iteritems()]
truths = [val == old_setup[key] for key, val in kwargs.items()]
if all(truths):
return True
else:
......@@ -2540,7 +2540,7 @@ class MCMCTransientSearch(MCMCSearch):
self.theta_keys = []
fixed_theta_dict = {}
for key, val in self.theta_prior.iteritems():
for key, val in self.theta_prior.items():
if type(val) is dict:
fixed_theta_dict[key] = 0
self.theta_keys.append(key)
......
......@@ -3,7 +3,7 @@
Provides functions to aid in calculating the optimal setup for zoom follow up
"""
from __future__ import division, absolute_import, print_function
import logging
import numpy as np
......
......@@ -31,7 +31,7 @@ def _optional_import ( modulename, shorthand=None ):
logging.debug('Successfully imported module %s%s.'
% (modulename, shorthandbit))
success = True
except ImportError, e:
except ImportError as e:
if e.message == 'No module named '+modulename:
logging.debug('No module {:s} found.'.format(modulename))
success = False
......@@ -111,7 +111,7 @@ def init_transient_fstat_map_features ( wantCuda=False, cudaDeviceName=None ):
' then checking all available devices...')
try:
context0 = pycuda.tools.make_default_context()
except pycuda._driver.LogicError, e:
except pycuda._driver.LogicError as e:
if e.message == 'cuDeviceGet failed: invalid device ordinal':
devn = int(os.environ['CUDA_DEVICE'])
raise RuntimeError('Requested CUDA device number {} exceeds' \
......
......@@ -315,10 +315,10 @@ class SemiCoherentGlitchSearch(Test):
Writer.tend = maxStartTime
FSB = Writer.predict_fstat()
print FSA, FSB
print(FSA, FSB)
predicted_FS = (FSA + FSB)
print(predicted_FS, FS)
print((predicted_FS, FS))
self.assertTrue(np.abs((FS - predicted_FS))/predicted_FS < 0.3)
......@@ -359,8 +359,8 @@ class MCMCSearch(Test):
search.run(create_plots=False)
_, FS = search.get_max_twoF()
print('Predicted twoF is {} while recovered is {}'.format(
predicted_FS, FS))
print(('Predicted twoF is {} while recovered is {}'.format(
predicted_FS, FS)))
self.assertTrue(
FS > predicted_FS or np.abs((FS-predicted_FS))/predicted_FS < 0.3)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment