def measure_phase(profile, template):
"""
measure_phase(profile, template):
Call FFTFIT on the profile and template to determine the
following parameters: shift,eshift,snr,esnr,b,errb,ngood
(returned as a tuple). These are defined as in Taylor's
talk at the Royal Society.
"""
c, amp, pha = fftfit.cprof(template)
pha1 = pha[0]
pha = Num.fmod(pha - Num.arange(1, len(pha) + 1) * pha1, TWOPI)
shift, eshift, snr, esnr, b, errb, ngood = fftfit.fftfit(profile, amp, pha)
return shift, eshift, snr, esnr, b, errb, ngood
from __future__ import print_function
#>>> print fftfit.__doc__
#This module 'fftfit' is auto-generated with f2py (version:2.13.175-1250).
#Functions:
# zbrent = zbrent(x1,x2,f1,f2,tol,tmp,pha,nsum)
# dchisqr = dchisqr(tau,tmp,r,nsum)
# cprof(y,c,amp,pha,nmax=len(y),nh=(len(c)-1))
# fccf(amp,pha,shift)
# ffft(d,npts,isign,ireal)
# fftfit(prof,s,phi,nmax,shift,eshift,snr,esnr,b,errb,ngood)
import numpy as num
from presto.psr_utils import gaussian_profile, TWOPI
from presto.fftfit import cprof, fftfit
template = gaussian_profile(64, 0.5, 0.1)
c, amp, pha = cprof(template)
#pha.savespace()
pha1 = pha[0]
pha = num.fmod(pha - num.arange(1, len(pha) + 1) * pha1, TWOPI)
for phs in [0.1, 0.3, 0.7]:
prof = gaussian_profile(64, phs, 0.1) + num.random.standard_normal(64)
shift, eshift, snr, esnr, b, errb, ngood = fftfit(prof, amp, pha)
print("True phs = %f, measured phs = %f +/- %f" %
(phs, shift / len(prof), eshift / len(prof)))