def test_fft_advance(input, expect):
computer = Fft(input)
computer.advance_n(100)
assert expect == computer.get_state()[:8]
audio = Audio(source={
'input': 'wav',
'path': 'resources/DaftPunk.wav',
'datasize': datasize
},
output=True)
if mode == 'mic':
audio = Audio(source={
'input': 'mic',
'datasize': datasize,
'rate': frate
},
output=False)
fft = Fft(datasize=datasize, frate=frate)
data = audio.sample_and_send()
fft.configure_fft(data)
fft.getDominantF()
fft.splitLevels()
fft.normalize_bin_values()
# bin_history.append(fft.stats['bin_values_normalized'][chan])
# print len(bin_history)
while (len(bin_history[0]) < history_length):
data = audio.sample_and_send()
fft.configure_fft(data)
fft.getDominantF()
def test_fft_get_pattern(dilate, length, expect):
p = Fft.get_pattern(dilate, length)
assert p == expect
def run(self):
print 'create client'
client = UdpClient(udp_ip=self.udp_ip,
udp_port_rec=self.udp_port_rec,
udp_port_send=self.udp_port_send)
client.connect()
if client.connected:
print 'client connected'
''' 1 - get source '''
datasize = 2048
frate = 44100
self.mode = 'mic'
if self.mode == 'wav':
audio = Audio(source={
'input': 'wav',
'path': 'resources/DaftPunk.wav',
'datasize': self.datasize
},
output=True)
if self.mode == 'mic':
audio = AudioStream(source={
'input': 'mic',
'datasize': self.datasize,
'rate': self.frate
},
output=False)
''' create fft '''
fft = Fft(datasize=datasize,
frate=frate,
gain=10e-4,
saturation_point=1024)
data = audio.sample_and_send()
fft.configure_fft(data)
fft.getDominantF()
fft.splitLevels()
fft.normalize_bin_values()
last_tick = time.time()
while True:
''' wait until next cycle '''
if (time.time() - last_tick) > self.process_period:
last_tick = time.time()
data = audio.sample_and_send()
fft.run_fft(data)
fft.getDominantF()
fft.splitLevels()
# fft.set_freq_bins_max()
fft.normalize_bin_values()
msg = ','.join(
[str(i) for i in fft.stats['bin_values_normalized']])
print msg
if not client.send(msg):
''' wait for reconnect '''
else:
time.sleep(.0001)
else:
print 'client not connected'
client.disconnect()
def __init__(self, spectype, cat_corr, ell=None, **kwargs):
'''
Class that describes power/bispectrum measurements
specify catalog, version, mock file number, file specifications (e.g. Nrandom),
fiber collision correction method, correction specifications (e.g. sigma, fpeak)
Parameters
----------
spectype : 'power' or 'bispec'
cat_corr : catalog and correction dictionary
'''
if spectype not in ['pk', 'bk']:
raise ValueError()
else:
self.type = spectype
if 'spec' not in cat_corr.keys():
if self.type == 'bk':
ell = 2 # this is a hack so that the FFT is from the quadrupole FFT
if ell is None:
raise ValueError("Specify ell (monopole: 0, quadrupole: 2, hexadecapole: 4)")
# default spectrum parameters
cat_corr['spec'] = {
'P0': 20000, #P0
'Lbox': 3600,
'Ngrid':360,
'ell': ell
}
self.ell = ell
if 'Ngrid' in kwargs.keys():
cat_corr['spec']['Ngrid'] = kwargs.pop('Ngrid')
else:
if self.type == 'bk':
cat_corr['spec']['ell'] = 2
elif 'ell' not in cat_corr['spec'].keys():
raise ValueError("Specify ell (monopole: 0, quadrupole: 2, hexadecapole: 4) in catcorr dictionary")
if ell is not None:
if ell != cat_corr['spec']['ell']:
raise ValueError
self.ell = cat_corr['spec']['ell']
self.cat_corr = cat_corr.copy()
self.kwargs = kwargs
if self.type == 'bk':
self.scale = np.float(self.cat_corr['spec']['Lbox'])
k_fund = (2.0*np.pi)/self.scale # k fundamental
self.k_fund = k_fund
try:
self.gal_data = CorrData(self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.rand_data = Random(self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.datafft = Fft('data', self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.randfft = Fft('random', self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
self.file_name = self.file()
class Spec(object):
def __init__(self, spectype, cat_corr, ell=None, **kwargs):
'''
Class that describes power/bispectrum measurements
specify catalog, version, mock file number, file specifications (e.g. Nrandom),
fiber collision correction method, correction specifications (e.g. sigma, fpeak)
Parameters
----------
spectype : 'power' or 'bispec'
cat_corr : catalog and correction dictionary
'''
if spectype not in ['pk', 'bk']:
raise ValueError()
else:
self.type = spectype
if 'spec' not in cat_corr.keys():
if self.type == 'bk':
ell = 2 # this is a hack so that the FFT is from the quadrupole FFT
if ell is None:
raise ValueError("Specify ell (monopole: 0, quadrupole: 2, hexadecapole: 4)")
# default spectrum parameters
cat_corr['spec'] = {
'P0': 20000, #P0
'Lbox': 3600,
'Ngrid':360,
'ell': ell
}
self.ell = ell
if 'Ngrid' in kwargs.keys():
cat_corr['spec']['Ngrid'] = kwargs.pop('Ngrid')
else:
if self.type == 'bk':
cat_corr['spec']['ell'] = 2
elif 'ell' not in cat_corr['spec'].keys():
raise ValueError("Specify ell (monopole: 0, quadrupole: 2, hexadecapole: 4) in catcorr dictionary")
if ell is not None:
if ell != cat_corr['spec']['ell']:
raise ValueError
self.ell = cat_corr['spec']['ell']
self.cat_corr = cat_corr.copy()
self.kwargs = kwargs
if self.type == 'bk':
self.scale = np.float(self.cat_corr['spec']['Lbox'])
k_fund = (2.0*np.pi)/self.scale # k fundamental
self.k_fund = k_fund
try:
self.gal_data = CorrData(self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.rand_data = Random(self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.datafft = Fft('data', self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
try:
self.randfft = Fft('random', self.cat_corr, **self.kwargs)
except NotImplementedError:
pass
self.file_name = self.file()
def read(self):
""" Read power/bispectrum of simulated/observed data catalog
"""
if self.type == 'pk': # power spectrum
if self.ell == 0: # monopole
col_index = [0, 1]
data_cols = ['k', 'p0k']
elif self.ell == 2: # quadrupoel
col_index = [0, 2, 1, 3]
data_cols = ['k', 'p2k', 'p0k', 'p4k']
elif self.ell == 4: # hexadecapole
col_index = [0, 3, 1, 2]
data_cols = ['k', 'p4k', 'p0k', 'p2k']
else:
raise NotImplementedError()
else: # bispectrum
# bispectrum v5 columns k1, k2, k3, P0(k1), P0(k2), P0(k3), B0, Q0, P2(k1), P2(k2), P2(k3), B2, Q2, dum, dum
col_index = [0, 1, 2, 3, 4, 5, 6, 7]
data_cols = ['k1', 'k2', 'k3', 'p0k1', 'p0k2', 'p0k3', 'bk', 'qk']
spec_data = np.loadtxt(
self.file_name,
unpack = True,
usecols = col_index
)
for i_col, col in enumerate(data_cols):
setattr(self, col, spec_data[i_col])
if self.type == 'bk':
self.k1 *= self.k_fund # k1 * k_fundamental to get h/Mpc
self.k2 *= self.k_fund
self.k3 *= self.k_fund
# some extra useful values
self.i_triangle = range(len(self.k1)) # triangle ID
self.avgk = (self.k1 + self.k2 + self.k3)/3.0 # average k
self.kmax = np.amax(np.vstack([self.k1, self.k2, self.k3]), axis=0) # max(k1,k2,k3)
return None
def file(self):
""" power/bispectrum file
"""
specdict = (self.cat_corr)['spec']
# powerspectrum or bispectrum
if self.type in ('pk'):
spec_str = 'POWER_'
elif self.type == 'bk':
spec_str = 'BISP_'
#if 'quad' not in specdict.keys():
# specdict['quad'] = False
#if specdict['quad']:
# spec_str += 'Q_'
if (self.type == 'pk') and (self.ell != 0):
spec_str += 'Q_'
self.data_file = self.gal_data.file_name
gal_file = (self.gal_data.file_name).split('/')[-1]
self.random_file = self.rand_data.file_name
spec_dir = direc('spec', self.cat_corr)
if self.type == 'pk':
specparam_str = ''.join([
'.grid', str(specdict['Ngrid']),
'.P0', str(specdict['P0']),
'.box', str(specdict['Lbox'])
])
elif self.type == 'bk':
specparam_str = ''.join([
'.grid', str(specdict['Ngrid']),
'.nmax40.ncut3.s3',
'.P0', str(specdict['P0']),
'.box', str(specdict['Lbox'])
])
else:
raise NotImplementedError
file_name = ''.join([
spec_dir,
spec_str,
gal_file,
specparam_str
])
return file_name
def build(self):
""" Calculate power/bispectrum of simulated/observed data catalog
"""
if 'clobber' not in (self.kwargs).keys():
bool_clobber = False
else:
bool_clobber = self.kwargs['clobber']
codeclass = Fcode(self.type, self.cat_corr)
spec_code = codeclass.code
spec_exe = codeclass.fexe()
# code and exe modification time
speccode_t_mod, specexe_t_mod = codeclass.mod_time()
if specexe_t_mod <= speccode_t_mod:
codeclass.compile()
# fft files
if not os.path.isfile(self.datafft.file_name) or bool_clobber:
self.datafft.build()
if not os.path.isfile(self.randfft.file_name):
self.randfft.build()
spec_cmd = codeclass.commandline_call(
datafft = self.datafft.file_name,
randfft = self.randfft.file_name,
powerfile = self.file_name
)
print spec_cmd
if any([not os.path.isfile(self.file_name), bool_clobber]):
print ''
print '-----------------------'
print 'Constructing '
print self.file_name
print '-----------------------'
print ''
print spec_cmd
print '-----------------------'
subprocess.call(spec_cmd.split())
else:
print ''
print '-----------------------'
print self.file_name
print 'Already Exists'
print '-----------------------'
print ''
return None
import os
from fft import Fft
if __name__ == "__main__":
this_dir = os.path.dirname(os.path.abspath(__file__))
input_path = os.path.join(this_dir, "input.txt")
with open(input_path) as f:
raw_code = f.readline()
fft = Fft(raw_code)
fft.advance_n(100)
print("Part 1")
print(f"{fft.get_state()}")
def build(self):
""" Calculate power/bispectrum of simulated/observed data catalog
"""
if 'clobber' not in (self.kwargs).keys():
bool_clobber = False
else:
bool_clobber = self.kwargs['clobber']
#if corrdict['name'] == 'fourier_tophat':
# if self.ell != 2:
# raise ValueError
# true_cat_corr = {
# 'catalog': catdict,
# 'correction': {'name': 'true'}
# }
# tr_gal = Data('data', true_cat_corr)
# fourier_tophat_Pk(self.cat_corr, self.file_name, tr_gal.file_name)
# return None
codeclass = Fcode(self.type, self.cat_corr)
spec_code = codeclass.code
spec_exe = codeclass.fexe()
# code and exe modification time
speccode_t_mod, specexe_t_mod = codeclass.mod_time()
if specexe_t_mod <= speccode_t_mod:
codeclass.compile()
# fft files
datafft = Fft('data', self.cat_corr, **self.kwargs)
if not os.path.isfile(datafft.file_name+'_0') or bool_clobber:
datafft.build()
randfft = Fft('random', self.cat_corr, **self.kwargs)
if not os.path.isfile(randfft.file_name+'_0'):
randfft.build()
spec_cmd = codeclass.commandline_call(
datafft = datafft.file_name,
randfft = randfft.file_name,
powerfile = self.file_name
)
print spec_cmd
if any([not os.path.isfile(self.file_name), bool_clobber]):
print ''
print '-----------------------'
print 'Constructing '
print self.file_name
print '-----------------------'
print ''
print spec_cmd
print '-----------------------'
subprocess.call(spec_cmd.split())
else:
print ''
print '-----------------------'
print self.file_name
print 'Already Exists'
print '-----------------------'
print ''
return None
