# coding: utf-8

# In[1]:

import matplotlib.pyplot as plt

from scipy.stats import binned_statistic as bn

from scipy.optimize import curve_fit as cf

from IPython.display import clear_output

import numpy as np

import os

from matplotlib.backends.backend_pdf import PdfPages

from PyAstronomy.pyasl import foldAt

from astropy.stats import LombScargle

from astropy import units as u

import shutil

get_ipython().run_line_magic('matplotlib', 'inline')

o=np.linspace(0,1,10000)

path = '../astars/'

path2 = path + 'Good/'

def curve(theta,a,phi):

return a*np.sin(2*np.pi*theta-(phi))+np.mean(cv)

# In[ ]:

for root, dirs, files in os.walk(path): #Walks through all the stars in a project directory

for file in files:

if (file == "comp.dat"):

ct = np.loadtxt(root+'/'+file)[0]#Loads composite light curve information

cv = np.loadtxt(root+'/'+file)[1]

t0 = np.min(ct)

if not os.path.isfile(root+'/pgram.dat'): #Computes Generalised L-S periodogram if it doesn't exist

frequency, power = LombScargle(ct*u.day, cv*u.mag).autopower(minimum_frequency=2/((np.max(ct)-np.min(ct))*u.day))

np.savetxt(root + '/pgram.dat',(power,frequency))

else: #If pgram.dat exists, this star has been looked at so skip

continue

P=foldAt(ct*u.day,1/frequency[np.argmax(power)]) #Phase-folds on day at maximum power

maxp = 1/frequency[np.argmax(power)]

pout,pcov=cf(curve,P,cv) #Least-squares fit sine wave to phase-folded LC

perr = np.sqrt(np.diag(pcov)) #Square root of diag(covariance) for errors

bins = bn(P,cv,statistic = 'median',bins = 15) #Sets up 15 bins to plot general form of pfolded LC

binsites = np.array([])

for i in range(len(bins[1]) - 1):

binsites = np.append(binsites, np.median([bins[1][i],bins[1][i+1]]))

f, ([ax1,ax2],[ax3,ax4]) = plt.subplots(2, 2, sharex=False,sharey = False,squeeze = True,figsize=(10,10))

ax1.set_ylabel("Brightness (mag)") #plots LC

ax1.set_xlabel("BJD - " + str(t0))

ax1.set_title("ASCC " + str(regex.findall(root)[0]))

ax1.invert_yaxis()

ax1.plot(ct-t0,cv,'black',marker='.',linestyle='None')

ax2.set_ylabel("Brightness (mag)") #Plots phase folded lc

ax2.set_xlabel("Phase")

ax2.set_title(str(np.around(maxp,4)) + " days, " + str(np.around(1/maxp,4)) + " c/d, " +str(np.around(np.abs(pout[0]),4)) +" mag")

ax2.invert_yaxis()

ax2.plot(P,cv,'gray',marker='.',linestyle='None')

ax2.plot(o,curve(o,*pout),'black',linewidth = 5.0)

ax2.plot(binsites,bins[0],'black',marker='o',markersize=7.0,linestyle='None')

ax2.set_xlim(0,1)

#ax1[0].axes.get_xaxis().set_visible(False)

ax3.set_ylabel("Normalized Power") #Plots time periodogram

ax3.set_xlabel("Time (Days)")

ax3.plot(1/frequency,power,'black')

ax3.tick_params(axis='both', which='minor')

ax4.set_ylabel("Normalized Power") #Plots frequency periodogram

ax4.set_xlabel("Frequency (c/d)")

ax4.plot(frequency,power,'black')

ax4.tick_params(axis='both', which='minor')

plt.show()

name = input("Good (l) or bad (else)?") #Initial sorting. Basically does it have a unique period to

#the distribution of systematics?

type(name)

if(name == 'l'):

clear_output()

pp=PdfPages(root + '/plots.pdf') #Saves copy of plots to folder

pp.savefig(f,bbox_inches='tight')

pp.close()

plt.close('all')

shutil.move(root,"../astars/Good") #Moves folder to step 2 (next code segment)

else:

clear_output()#Moves onto the next star

plt.close("all")

# In[ ]:

for root, dirs, files in os.walk(path): #Individual sorting code

print(root)

for file in files:

name = 'q' #Sets intiial variables for a given star in the "Good" folder of a project. name is what is used

#to process what to do next.

d = True

in6 = 'f'

while(d == True):

clear_output()

plt.close("all")

if (file == "comp.dat"):

ct = np.loadtxt(root+'/'+file)[0]

cv = np.loadtxt(root+'/'+file)[1]

t0 = np.min(ct)

power,frequency = np.loadtxt(root+'/pgram.dat')

if name != 'c':

maxp = 1/frequency[np.argmax(power)]

else:

maxp = in5

P=foldAt(ct*u.day,maxp) #Recomputes the phas-fold as needed

pout,pcov=cf(curve,P,cv)

perr = np.sqrt(np.diag(pcov))

bins = bn(P,cv,statistic = 'median',bins = 40)

binsites = np.array([])

for i in range(len(bins[1]) - 1):

binsites = np.append(binsites, np.median([bins[1][i],bins[1][i+1]]))

f, ([ax1,ax2],[ax3,ax4]) = plt.subplots(2, 2, sharex=False,sharey = False,squeeze = True,figsize=(10,10))

#Draws same plot as above, but with possible new options (below).

ax1.set_ylabel("Brightness (mag)")

ax1.set_xlabel("BJD - " + str(t0))

ax1.set_title("ASCC " + str(regex.findall(root)[0]))

if name == 'z':

ax1.set_xlim(in1,in2)

ax1.invert_yaxis()

ax1.plot(ct-t0,cv,'black',marker='.',linestyle='None')

ax2.set_ylabel("Brightness (mag)")

ax2.set_xlabel("Phase")

ax2.set_title(str(np.around(maxp,4)) + " days, " + str(np.around(1/maxp,4)) + " c/d, " +str(np.around(np.abs(pout[0]),4)) +" mag")

ax2.invert_yaxis()

ax2.plot(P,cv,'gray',marker='.',linestyle='None')

if not (in6 == 'y'):

ax2.plot(o,curve(o,*pout),'black',linewidth = 5.0)

ax2.plot(binsites,bins[0],'black',marker='o',markersize=7.0,linestyle='None')

ax2.set_xlim(0,1)

ax3.set_ylabel("Normalized Power")

ax3.set_xlabel("Time (Days)")

ax3.plot(1/frequency,power,'black')

if name == 'x':

ax3.set_xlim(in3,in4)

ax3.tick_params(axis='both', which='minor')

ax4.set_ylabel("Normalized Power")

ax4.set_xlabel("Frequency (c/d)")

ax4.plot(frequency,power,'black')

if name == 'x':

if(in3 == 0):

ax4.set_xlim(0,1/in4)

else:

ax4.set_xlim(1/in3,1/in4)

ax4.tick_params(axis='both', which='minor')

plt.show()

if name =='r':

print(newmax)

name = input("z = zoom light curve, x = zoom pgram, c = phase-fold d, s = save plot, v= remove,\n f = cepheid, g = EB, r = search, \np=DS, y = LV, i = rot, u = irregular, j = Be, else=next")

type(name)

if(name == 'z'): #Redraw with new LC time limits

in1 = float(input("Lower limit = "))

type(in1)

in2 = float(input("Upper limit = "))

type(in2)

elif(name == 'x'): #Redraw pgrams with new time/frequency bounds

in3 = float(input("Lower limit = "))

type(in3)

in4 = float(input("Upper limit = "))

type(in4)

elif(name == 'c'): #Phase-fold on new period

in5 = float(input("New Period = "))

type(in5)

in6 = input("Remove Sinusoid (y)?")

type(in6)

elif(name == 'r'): #Searches for the stronges period in the given range of days.

in7 = float(input("Lower limit = "))

type(in7)

in8 = input("Upper limt = ")

type(in8)

days2 = 1/frequency[np.where(np.logical_and(1/frequency>=float(in7),1/frequency<=float(in8)))]

power2 = power[np.where(np.logical_and(1/frequency>=float(in7),1/frequency<=float(in8)))]

newmax = days2[np.argmax(power2)]

elif(name == 's'): #Saves current plot over plots file in folder

pp=PdfPages(root + '/plots.pdf')

pp.savefig(f,bbox_inches='tight')

pp.close()

elif(name == 'v'): #Each of these sorts into respective variable type.

shutil.move(root,path)

d = False

elif(name == 'f'):

shutil.move(root,'../Cepheids')

d = False

elif(name == 'y'):

shutil.move(root,'../Long_Variable')

d = False

elif(name == 'p'):

shutil.move(root,'../DS')

d = False

elif(name == 'u'):

shutil.move(root,'../Irregular')

d = False

elif(name == 'i'):

shutil.move(root,'../Rot')

d = False

elif(name == 'g'):

shutil.move(root,'../EBs')

d = False

elif(name == 'j'):

shutil.move(root,'../BE')

d = False

else:

d = False