-
Notifications
You must be signed in to change notification settings - Fork 0
/
phase_dependent_amplitude.py
367 lines (309 loc) · 12 KB
/
phase_dependent_amplitude.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
import numpy as np
import matplotlib.pyplot as plt
import lightkurve as lk
from smurfs import Smurfs
from tqdm import tqdm, trange
import multiprocessing as mp
import ast
import os
from scipy.optimize import curve_fit
from lmfit import Model
class mode():
def __init__(self, f, amp, phase, snr, label = ""):
self.f = f
self.amp = amp
self.phase = phase
self.snr = snr
self._label = label
@property
def label(self) -> str:
"""
Returns the label of the found frequency
"""
return self._label
def pda(time, amp, frequency, phase, phase0 = 0.3, sigma = 0.05, defect = 0.5):
period = 1.66987725
phases = time%period/period
return amp*(1- defect*np.exp(-(phases - phase0) ** 2 / (2 * sigma ** 2)))*np.sin(2. * np.pi * (frequency * time + phase))
def get_result(list, dict):
freq = 1/1.66987725
f0 = list[0].f.n
used = []
for f in list:
num = int(np.round((f.f.n-f0)/freq, 0))
if num in used:
continue
else:
used.append(num)
try:
if num != 0:
dict[f"amp_{num}"].append(f.amp.n/list[0].amp.n)
dict[f"phase_{num}"].append(f.phase.n-list[0].phase.n)
else:
dict[f"amp_{num}"].append(f.amp.n)
dict[f"phase_{num}"].append(f.phase.n)
except:
continue
for i in range(-5, 6):
if i not in used:
dict[f"amp_{i}"].append(0)
dict[f"phase_{i}"].append(0)
return dict
def simulate(run, time, folder, defect):
np.random.seed(int.from_bytes(os.urandom(4), byteorder='little'))
result = {}
for i in range(-5, 6):
result[f"amp_{i}"] = []
result[f"phase_{i}"] = []
for _ in trange(200, desc = f'Run {run+1}', position = run+1):
flux = pda(time, np.random.uniform(0.001, 0.003,1), np.random.uniform(10, 20, 1), np.random.uniform(0,1,1), defect = defect)\
+ np.random.normal(0, 0.000165, len(time))
s = Smurfs(time = time, flux = flux, quiet_flag=True)
s.run(improve_fit=True, f_max = 40, mode='scipy')
#s.improve_result()
res = s.result["f_obj"].values
result = get_result(res, result)
f = open(f'pda_simulations/{folder}/' + f'{run}.txt', "w")
f.write(str(result))
f.close()
def simulate_multiple(folder, defect= 0.5):
n = 5
if not os.path.exists('pda_simulations/'+str(folder)):
os.mkdir('pda_simulations/'+str(folder))
pool = mp.Pool(processes=n)
data = np.loadtxt('endurance/' + 'Final_Pulsation_LC.txt').T
time = data[0][:20000]
runs = []
for i in range(n):
runs.append((i, time, folder, defect))
for _ in pool.starmap(simulate, runs):
pass
def load_dict(folder, number):
file = folder + f'{number}.txt'
with open(file, 'r') as file:
data = file.read().replace('\n', '')
return ast.literal_eval(data)
def load_all(folder):
result = {}
for i in range(-5, 6):
result[f"amp_{i}"] = []
result[f"phase_{i}"] = []
for i in range(0,5):
step = load_dict(f'pda_simulations/{folder}/', i)
for key in step.keys():
result[key].extend(step[key])
return result
def plot_simulations(folder, ax_amp, ax_phase, fc = 'gray'):#Initiate two figures acc to fig_phase, ax_phase = plt.subplots(3, 4, figsize=(12,7))
result = load_all(folder)
amp_key = [f"amp_{i}" for i in range(-5, 6)]
phase_key = [f"phase_{i}" for i in range(-5, 6)]
filterByKey = lambda keys: {x: result[x] for x in keys}
count = 0
for key in filterByKey(amp_key):
data = np.array(result[key])
data = data[data != 0]
ax = ax_amp[int(count/4)][count%4]
if key[-1] == "0":
ax.hist(np.array(data), bins = 50, edgecolor = 'k', facecolor = fc, label = key, alpha = 0.5)
ax.legend(fontsize = 'x-small')
ax.set_xlabel('$A_0$')
ax.set_ylabel('occurence')
else:
ax.hist(np.array(data), bins = 50, edgecolor = 'k', facecolor = fc, label = key, alpha = 0.5)
ax.legend(fontsize = 'x-small')
ax.set_xlabel('$A$/$A_0$')
ax.set_ylabel('occurence')
count+=1
count = 0
for key in filterByKey(phase_key):
data = np.array(result[key])
data = data[data != 0]
ax = ax_phase[int(count/4)][count%4]
ax.hist((np.array(data)+1)%1, bins = 50, edgecolor = 'k', facecolor = fc, label = key, alpha = 0.5)
ax.legend(fontsize = 'x-small')
ax.set_xlabel('phase')
ax.set_ylabel('occurence')
count += 1
#fig_amp.tight_layout(h_pad=0, w_pad=0)
#fig_phase.tight_layout(h_pad=0, w_pad=0)
#plt.show()
#plt.close()
return ax_amp, ax_phase
class run():
def __init__(self, num, time):
self.n = num
self.time = time
def plot_mult_sims(folders, colors):
fig_amp, ax_amp = plt.subplots(3, 4, figsize=(12,7))
fig_phase, ax_phase = plt.subplots(3, 4, figsize=(12,7))
for f, c in zip(folders, colors):
ax_amp, ax_phase = plot_simulations(f, ax_amp, ax_phase, fc = c)
plt.show()
def gauss(x, a, x0, sigma):
return a * np.exp(-(x - x0) ** 2 / (2 * sigma ** 2))
def line(x, a, ):
return a * x
def square(x, a, b):
return a * x + b*x**2
def get_amplitude_value(folder, key):
result = load_all(folder)
data = np.array(result[key])
data = data[data != 0]
vals, edges =np.histogram(data, 30)
xs = (edges[1:]+edges[:-1])/2
try:
popt, pcov = curve_fit(gauss, xs, vals, p0=[100, np.mean(data), np.std(data)])
return popt[1], popt[2], len(data)
except RuntimeError:
errfig, errax = plt.subplots(figsize = (12,7))
x = np.linspace(np.min(data), np.max(data), 500)
errax.plot(x, gauss(x, *[100, np.mean(data), np.std(data)]), 'r-')
errax.bar(xs, vals, width = edges[1]-edges[0], edgecolor = 'k', facecolor = 'gray')
errax.set_title(key)
return 0, 0, 0
def plot_amplitude_statistics(figsize, dpi):
c = ['#FF5733', '#FFB900', '#93FF00', '#019539', '#00D1EA', '#000BEA', '#BF00EA', '#806979', '#000000', '#3e197c', '#ffb600']
fig, ax = plt.subplots(figsize = figsize, dpi = dpi)
amp_key = [f"amp_{i}" for i in range(-5, 6)]
count = 0
for key in range(-5, 6):
if key != 0:
means = []
sds = []
defects = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8]
for folder in range(4, 12):
m, s, length = get_amplitude_value(folder, f'amp_{key}')
means.append(m)
sds.append(s)
marker = '^'
if key > 0:
marker = 's'
#popt, pcov = curve_fit(square, defects, means, sigma=sds, p0=[(means[-1]-means[0])/0.7, 0], absolute_sigma=True)
popt, pcov = curve_fit(line, defects, means, sigma=sds, p0=[(means[-1]-means[0])/0.7])#, absolute_sigma=True)
x = np.linspace(0,0.8, 500)
print(f"$A_{ {key} }$")
N = np.array(means)
#Nexp = square(np.array(defects), *popt)
Nexp = line(np.array(defects), *popt)
r = N - Nexp
chisq = np.sum((r / np.array(sds)) ** 2)
df = len(means) - len(popt)
print("chisq =", chisq, "df =", df, "reduced chisq", chisq/df, 'num= ', length)
print('###################')
ax.errorbar(defects, means, yerr = sds, color = c[key+5], marker = marker, markerfacecolor= 'none',
markeredgewidth = 1, capsize = 2, lw =0, elinewidth = 1, label = f"$\\tilde{{A}}_{ {key} }$")# $"+ "$\\frac{\chi^2}{DoF}=$" +f"{np.round(chisq/df, 3)}")
#ax.plot(x, square(x, *popt), color = c[count])
lw ='-'
if key > 0:
lw ='--'
ax.plot(x, line(x, *popt), color = c[key+5], ls = lw)
count+=1
ax.legend(fontsize = 'small', ncol = 4, labelspacing=0.05, handletextpad = 0.1)
ax.set_xlabel('Defect')
ax.set_ylabel('relative Amplitude')
# plt.tight_layout()
#plt.savefig('pda_simulations/plots/line.png', bbox_inches = 0)
# plt.show()
return fig, ax
def sin(x: np.ndarray, amp: float, f: float, phase: float) -> np.ndarray:
return amp * np.sin(2. * np.pi * (f * x + phase))
def _lmfit_fit(result, time, flux):
models = []
for f in result:
m = Model(sin, prefix=f._label)
m.set_param_hint(f.label + 'amp', value=f.amp, min=0 * f.amp,
max=2 * f.amp)
m.set_param_hint(f.label + 'f', value=f.f, min=0.8*f.f,
max=1.2 * f.f)
m.set_param_hint(f.label + 'phase', value=f.phase, min=0,
max=1)
models.append(m)
model: Model = np.sum(models)
fit_result = model.fit(flux, x=time)
for f in result:
#sigma_amp, sigma_f, sigma_phi = m_od_uncertainty(self.lc, fit_result.values[f._label + 'amp'])
f.amp = fit_result.values[f.label + 'amp']
f.f = fit_result.values[f.label + 'f']
f.phase =fit_result.values[f.label + 'phase']
return result
def simulate_fit(run, time, folder, defect):
np.random.seed(int.from_bytes(os.urandom(4), byteorder='little'))
result_sim = {}
f_orb = 1/1.66987725
result_sim[f"amp"] = []
result_sim[f"phase"] = []
result_sim[f"freq"] = []
for i in range(-5, 6):
result_sim[f"amp_{i}"] = []
result_sim[f"phase_{i}"] = []
for _ in trange(200, desc = f'Run {run+1}', position = run+1):
amp = np.random.uniform(0.001, 0.003,1)[0]
freq = np.random.uniform(10, 20, 1)[0]
phi = np.random.uniform(0,1,1)[0]
result_sim[f"amp"].append(amp)
result_sim[f"phase"].append(phi)
result_sim[f"freq"].append(freq)
flux = pda(time, amp, freq, phi, defect = defect)\
+ np.random.normal(0, 0.000165, len(time))
result = []
for i in range(-5,6):
if i ==0:
result.append(mode(freq + i* f_orb, amp, phi, 10, label="F"+str(i).replace('-', '_')))
else:
result.append(mode(freq + i * f_orb, amp*0.05, phi, 10, label="F" + str(i).replace('-', '_')))
result = _lmfit_fit(result, time, flux)
for f in result:
f._label = f._label.replace('_', '-')[1:]
if f._label == '0':
f0 = f
for f in result:
num = int(f._label)
if num != 0:
result_sim[f"amp_{num}"].append(f.amp / f0.amp)
result_sim[f"phase_{num}"].append(f.phase - f0.phase)
else:
result_sim[f"amp_{num}"].append(f.amp)
result_sim[f"phase_{num}"].append(f.phase)
f = open(f'pda_simulations/fit/{folder}/' + f'{run}.txt', "w")
f.write(str(result_sim))
f.close()
def simulate_multiple_fit(folder, defect= 0.5):
n = 5
if not os.path.exists('pda_simulations/fit/'+str(folder)):
os.mkdir('pda_simulations/fit/'+str(folder))
pool = mp.Pool(processes=n)
data = np.loadtxt('endurance/' + 'Final_Pulsation_LC.txt').T
time = data[0][:20000]
runs = []
for i in range(n):
runs.append((i, time, folder, defect))
for _ in pool.starmap(simulate_fit, runs):
pass
def get_phase_tda(folder):
tda_input = []
result = load_all(folder)
amps = result['amp_0']
ph_keys = []
for i in range(-5, 6):
if i !=0:
ph_keys.append(f"phase_{i}")
num = len(amps)
for i in range(num):
vert = []
for key in ph_keys:
vert.append((1+result[key][i])%1)
if 0 in vert:
print(vert)
else:
tda_input.append(vert)
return np.array(tda_input)
'''
f8 = get_phase_tda(8)
f7 = get_phase_tda(6)
from ripser import ripser
from persim import plot_diagrams
dgm8 = ripser(np.array(f8))['dgms'][1]
dgm7 = ripser(np.array(f7))['dgms'][1]
#print(dgm7, dgm8)
plot_diagrams([dgm8, dgm7] , labels=['Clean $H_1$', 'Noisy $H_1$'], show = True)
'''