def info():
"""Print date and version information as a html-table."""
# Define styles
style1 = " style='border: 2px solid #fff; text-align: left;'"
style2 = " style='background-color: #ccc;"
style2 += "width: 10%; border: 2px solid #fff;'"
# Print date and time info as title
html = "<h3>%s</h3>" % time.strftime('%a %b %d %H:%M:%S %Y %Z')
# Start table
html += '<table>'
# empymod, OS, threads, IPython
html += "<tr" + style1 + ">"
html += "<td" + style2 + ">%s</td>" % empymod.__version__
html += "<td" + style1 + ">empymod</td>"
html += "<td" + style2 + ">%s</td>" % platform.system()
html += "<td" + style1 + ">OS</td>"
html += "<td" + style2 + ">%s</td>" % numexpr.detect_number_of_threads()
html += "<td" + style1 + ">CPU(s)</td>"
html += "<td" + style2 + ">%s</td>" % IPython.__version__
html += "<td" + style1 + ">IPython</td>"
html += "</tr>"
# numpy, scipy, numexpr, matplotlib
html += "<tr" + style1 + ">"
html += "<td" + style2 + ">%s</td>" % numpy.__version__
html += "<td" + style1 + ">numpy</td>"
html += "<td" + style2 + ">%s</td>" % scipy.__version__
html += "<td" + style1 + ">scipy</td>"
html += "<td" + style2 + ">%s</td>" % numexpr.__version__
html += "<td" + style1 + ">numexpr</td>"
html += "<td" + style2 + ">%s</td>" % matplotlib.__version__
html += "<td" + style1 + ">matplotlib</td>"
html += "</tr>"
# sys.version
html += "<tr" + style1 + ">"
html += "<td" + style1 + " colspan='8'>%s</td>" % sys.version
html += "</tr>"
# vml version
html += "<tr" + style2 + ">"
html += "<td" + style2 + " colspan='8'>%s</td>" % numexpr.get_vml_version()
html += "</tr>"
# Finish table
html += "</table>"
return HTML(html)
def test_omp_num_threads(self):
with _environment('OMP_NUM_THREADS', '5'):
self.assertEquals(5, numexpr.detect_number_of_threads())
def __init__(self, xsrc, ysrc, sch1, sch2, **kwargs):
"""Kappa analysis of a shear map.
:param list xsrc: x coordinates
:param list xsrc: y coordinates
:param list sch1: schear 1
:param list sch2: schear 2
:param list filt: filter to apply
Possible kwargs:
:param int step: Step of the grid
:param float rinner: Inner cut in pixel
:param float router: Outer cut in pixel
:param float theta0: Maturi outer cut in pixel
:param float aprad: Aperture mass outer cut in pixel
"""
assert len(xsrc) == len(ysrc) == len(sch1) == len(sch2)
# numba?
self.use_numba = kwargs.get("numba", False) and 'numba' in sys.modules
# numexpr?
self.use_numexpr = kwargs.get("numexpr", False) and 'numexpr' in sys.modules
if self.use_numba:
for func in to_jit:
apply_decorator(func, numba.jit)
for func in to_vectorize:
apply_decorator(func, numba.vectorize)
elif self.use_numexpr:
numexpr.set_num_threads(numexpr.detect_number_of_threads())
# Make sure all list are actually numpy arrays
xsrc, ysrc, sch1, sch2 = [np.array(x.tolist()) for x in [
xsrc, ysrc, sch1, sch2]]
# Do we have to filter?
if kwargs.get('filt', None) is not None:
assert len(kwargs['filt']) == len(xsrc)
self._idata = {'xsrc': xsrc, 'ysrc': ysrc,
'sch1': sch1, 'sch2': sch2,
'flag': np.array(kwargs.get('filt'))}
print("Filtering data")
print(" - Before cut: %i sources" % len(xsrc))
xsrc, ysrc, sch1, sch2 = [arr[np.array(kwargs.get('filt'))]
for arr in [xsrc, ysrc, sch1, sch2]]
print(" - After cut: %i sources" % len(xsrc))
self.filtered = True
else:
self._idata = None
self.filtered = False
# Save the data
self.data = {'xsrc': xsrc, 'ysrc': ysrc,
'sch1': sch1, 'sch2': sch2}
# Get or compute some useful parameters
self.parameters = {
# size of the image
'sizex': max(xsrc) - min(xsrc),
'sizey': max(ysrc) - min(ysrc),
# step, useful for test purpose
'step': kwargs.get('step', 1),
# number of points for each axis
'nxpoints': int((max(xsrc) - min(xsrc)) / kwargs.get('step', 1)),
'nypoints': int((max(ysrc) - min(ysrc)) / kwargs.get('step', 1)),
# inner and outer radius
'rinner': kwargs.get('rinner', 500.0),
'router': kwargs.get('router', 8000.0),
# maximum radius
'rmax': np.sqrt((max(xsrc) - min(xsrc))**2 + (max(ysrc) - min(ysrc))**2),
'theta0': kwargs.get('theta0', 6000.0),
'aprad': kwargs.get('aprad', 6000.0)
}
# Define needed dictionnary
self.maps = self.weights = {}
# Get the weights and the maps
self._get_weights()
self._get_kappa(xsampling=kwargs.get('xsampling', 10))
self.save_maps()
def test_numexpr_num_threads(self):
with _environment("OMP_NUM_THREADS", "5"):
with _environment("NUMEXPR_NUM_THREADS", "3"):
self.assertEquals(3, numexpr.detect_number_of_threads())
def __init__(self, xsrc, ysrc, sch1, sch2, **kwargs):
"""Kappa analysis of a shear map.
:param list xsrc: x coordinates
:param list xsrc: y coordinates
:param list sch1: schear 1
:param list sch2: schear 2
:param list filt: filter to apply
Possible kwargs:
:param int step: Step of the grid
:param float rinner: Inner cut in pixel
:param float router: Outer cut in pixel
:param float theta0: Maturi outer cut in pixel
:param float aprad: Aperture mass outer cut in pixel
"""
assert len(xsrc) == len(ysrc) == len(sch1) == len(sch2)
# numba?
self.use_numba = kwargs.get("numba", False) and 'numba' in sys.modules
# numexpr?
self.use_numexpr = kwargs.get("numexpr", False) and 'numexpr' in sys.modules
if self.use_numba:
for func in to_jit:
apply_decorator(func, numba.jit)
for func in to_vectorize:
apply_decorator(func, numba.vectorize)
elif self.use_numexpr:
numexpr.set_num_threads(numexpr.detect_number_of_threads())
# Make sure all list are actually numpy arrays
xsrc, ysrc, sch1, sch2 = [np.array(x.tolist()) for x in [
xsrc, ysrc, sch1, sch2]]
# Do we have to filter?
if kwargs.get('filt', None) is not None:
assert len(kwargs['filt']) == len(xsrc)
self._idata = {'xsrc': xsrc, 'ysrc': ysrc,
'sch1': sch1, 'sch2': sch2,
'flag': np.array(kwargs.get('filt'))}
print("Filtering data")
print(" - Before cut: %i sources" % len(xsrc))
xsrc, ysrc, sch1, sch2 = [arr[np.array(kwargs.get('filt'))]
for arr in [xsrc, ysrc, sch1, sch2]]
print(" - After cut: %i sources" % len(xsrc))
self.filtered = True
else:
self._idata = None
self.filtered = False
# Save the data
self.data = {'xsrc': xsrc, 'ysrc': ysrc,
'sch1': sch1, 'sch2': sch2}
# Get or compute some useful parameters
self.parameters = {
# size of the image
'sizex': max(xsrc) - min(xsrc),
'sizey': max(ysrc) - min(ysrc),
# step, useful for test purpose
'step': kwargs.get('step', 1),
# number of points for each axis
'nxpoints': int((max(xsrc) - min(xsrc)) / kwargs.get('step', 1)),
'nypoints': int((max(ysrc) - min(ysrc)) / kwargs.get('step', 1)),
# inner and outer radius
'rinner': kwargs.get('rinner', 500.0),
'router': kwargs.get('router', 8000.0),
# maximum radius
'rmax': np.sqrt((max(xsrc) - min(xsrc))**2 + (max(ysrc) - min(ysrc))**2),
'theta0': kwargs.get('theta0', 6000.0),
'aprad': kwargs.get('aprad', 6000.0)
}
# Define needed dictionnary
self.maps = self.weights = {}
# Get the weights and the maps
self._get_weights()
self._get_kappa(xsampling=kwargs.get('xsampling', 10))
self.save_maps()
def test_numexpr_num_threads(self):
with _environment('OMP_NUM_THREADS', '5'):
with _environment('NUMEXPR_NUM_THREADS', '3'):
self.assertEqual(3, numexpr.detect_number_of_threads())
from math import nan
matplotlib.use('Agg')
from matplotlib import pyplot
#from VTKWrapper import saveToVTK
import os
from timeit import default_timer as timer
#from numba import jit
from functions import sumf, equ, ucprod, copyfunc, allfunc
from functions import set_omega
from concurrent.futures import ThreadPoolExecutor
tstart = timer()
from numpy import *
#os.system("taskset -p 0xff %d" % os.getpid())
#print('number of cores detected is '); print(ne.detect_number_of_threads())
ne.set_num_threads(ne.detect_number_of_threads())
# Plot Settings
Pinterval = 1000 # iterations to print next output
SavePlot = False #save plots
SaveVTK = False # save files in vtk format
project = 'ldc' #for naming output files
OutputFolder = './output'
CurrentFolder = os.getcwd()
# Lattice Parameters
maxIt = 3000 # time iterations
Re = 1000.0 # Reynolds number 100 400 1000 3200 5000 7500 10000
#Number of cells
xsize, ysize = 320, 320
def good_process_number(total_cores, numexpr=True):
if numexpr:
thread_num = ne.detect_number_of_threads()
else:
thread_num = 1
return max(1, total_cores // thread_num)
