single_analysis.py

#!/usr/bin/env python3
import os; os.environ['OMP_NUM_THREADS']='1'
from analysis_tools_cython import *
from functools import reduce
import sys
import matplotlib.pyplot as plt
import argparse

parser = argparse.ArgumentParser(description='Analyse target lightcurve.')
parser.add_argument(help='Target lightcurve file',nargs=1,dest='fits_file')
parser.add_argument('-n', help='No graphical output', action='store_true')
parser.add_argument('-q', help='Keep only points with SAP_QUALITY=0',action='store_true')

args = parser.parse_args()

table=import_lightcurve(args.fits_file[0], args.q)

timestep = calculate_timestep(table)
t,flux,quality,real = clean_data(table)
N = len(t)
ones = np.ones(N)

flux = normalise_flux(flux)

filteredflux = fourier_filter(flux,8)
A_mag = np.abs(np.fft.rfft(flux))
periodicnoise = flux-filteredflux
sigma = flux.std()

flux_ls = np.copy(flux)
lombscargle_filter(t,flux_ls,real,0.05)
periodicnoise_ls = flux - flux_ls
flux_ls = flux_ls * real

T1 = test_statistic_array(filteredflux,60)
T = test_statistic_array(flux_ls,60)
data = nonzero(T)

# Find minimum test statistic value, and its location.
m,n = np.unravel_index(T.argmin(),T.shape)
minT = T[m,n]
minT_time = t[n]
minT_duration = m*timestep
print("Maximum transit chance:")
print("   Time =",round(minT_time,2),"days.")
print("   Duration =",round(minT_duration,2),"days.")
print("   T =",round(minT,1))
print("   T/sigma =",round(minT/data.std(),1))

trans_start = n - math.floor((m-1)/2)
trans_end = trans_start + m
print("Transit depth =",round(flux[trans_start:trans_end].mean(),6))

# Transit shape calculation
if n-3*m >= 0 and n+3*m < N:
    t2 = t[n-3*m:n+3*m]
    x2 = flux_ls[n-3*m:n+3*m]
    q2 = quality[n-3*m:n+3*m]
    background = (sum(x2[:1*m]) + sum(x2[5*m:]))/(2*m)
    x2 -= background
    paramsgauss = single_gaussian_curve_fit(t2,-x2)
    y2 = -gauss(t2,*paramsgauss)
    paramscomet = comet_curve_fit(t2,-x2)
    w2 = -comet_curve(t2,*paramscomet)

    scores = [score_fit(x2,fit) for fit in [y2, w2]]
    print(scores)
    print("Asym score:",round(scores[0]/scores[1],4))

    qual_flags = reduce(lambda a,b: a|b, q2)
    print("Quality flags:",qual_flags)

# Classify events
asym,_,_=calc_shape(m,n,t,flux)
print(classify(m,n,real,asym))

# Skip plotting if no graphical output set
if args.n:
    sys.exit()


#plt.xkcd()
fig1,axarr = plt.subplots(4)
axarr[0].plot(A_mag)
axarr[1].plot(t,flux+ones,t,periodicnoise_ls+ones)
axarr[2].plot(t,flux_ls+ones)
cax = axarr[3].imshow(T)
axarr[3].set_aspect('auto')
fig1.colorbar(cax)

#params = double_gaussian_curve_fit(T)
fig2 = plt.figure()
ax2 = fig2.add_subplot(111)
#T_test_nonzero = np.array(data)
#_,bins,_ = ax2.hist(T_test_nonzero,bins=100,log=True)
#y = np.maximum(bimodal(bins,*params),10)
#ax2.plot(bins,y)
try:
    ax2.plot(t2,x2,t2,y2,t2,w2)
except:
    pass

plt.show()