-
Notifications
You must be signed in to change notification settings - Fork 0
/
filaments_model_1h.py
executable file
·311 lines (220 loc) · 11.2 KB
/
filaments_model_1h.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
import os, yaml, argparse, sys, logging , pyfits, emcee, tabletools, cosmology, filaments_tools, nfw, plotstools, filament
import numpy as np
import pylab as pl
import warnings
warnings.simplefilter('once')
logger = logging.getLogger("fil_model_1h")
logger.setLevel(logging.INFO)
log_formatter = logging.Formatter("%(asctime)s %(name)s %(levelname)s %(message)s ","%Y-%m-%d %H:%M:%S")
stream_handler = logging.StreamHandler(sys.stdout)
stream_handler.setFormatter(log_formatter)
logger.addHandler(stream_handler)
logger.propagate = False
redshift_offset = 0.2
weak_limit = 0.2
class modelfit():
def __init__(self):
self.grid_z_edges = np.linspace(0,2,10);
self.grid_z_centers = plotstools.get_bins_centers(self.grid_z_edges)
self.prob_z = None
self.sigma_g = 0.2
self.inv_sq_sigma_g = None
self.n_model_evals = 0
self.sampler = None
self.n_samples = 1000
self.n_walkers = 10
self.n_grid = 10
self.save_all_models = False
self.n_dim = 1
self.parameters = [None]*self.n_dim
self.parameters[0] = {}
self.parameters[0]['name'] = 'halo_M200'
self.parameters[0]['box'] = {'min' : 13, 'max': 15}
self.shear_g1 = None
self.shear_g2 = None
self.shear_n_gals = None
self.shear_u_arcmin = None
self.shear_v_arcmin = None
self.halo_u_arcmin = None
self.halo_v_arcmin = None
def plot_shears(self,g1,g2,limit_mask=None,unit='arcmin',quiver_scale=1):
ephi=0.5*np.arctan2(g2,g1)
if limit_mask==None:
emag=np.sqrt(g1**2+g2**2)
else:
emag=np.sqrt(g1**2+g2**2) * limit_mask
nuse=1
line_width=0.005* quiver_scale
if unit=='arcmin':
pl.quiver(self.shear_u_arcmin[::nuse],self.shear_v_arcmin[::nuse],emag[::nuse]*np.cos(ephi)[::nuse],emag[::nuse]*np.sin(ephi)[::nuse],linewidths=0.001,headwidth=0., headlength=0., headaxislength=0., pivot='mid',color='r',label='original',scale=quiver_scale , width = line_width)
pl.xlim([min(self.shear_u_arcmin),max(self.shear_u_arcmin)])
pl.ylim([min(self.shear_v_arcmin),max(self.shear_v_arcmin)])
elif unit=='Mpc':
pl.quiver(self.shear_u_mpc[::nuse],self.shear_v_mpc[::nuse],emag[::nuse]*np.cos(ephi)[::nuse],emag[::nuse]*np.sin(ephi)[::nuse],linewidths=0.001,headwidth=0., headlength=0., headaxislength=0., pivot='mid',color='r',label='original',scale=quiver_scale , width = line_width)
pl.xlim([min(self.shear_u_mpc),max(self.shear_u_mpc)])
pl.ylim([min(self.shear_v_mpc),max(self.shear_v_mpc)])
pl.axis('equal')
def plot_shears_mag(self,g1,g2):
emag=np.sqrt(g1**2+g2**2)
pl.scatter(self.shear_u_arcmin,self.shear_v_arcmin,50,c=emag)
pl.colorbar()
pl.xlim([min(self.shear_u_arcmin),max(self.shear_u_arcmin)])
pl.ylim([min(self.shear_v_arcmin),max(self.shear_v_arcmin)])
pl.axis('equal')
def plot_residual_whisker(self,model_g1,model_g2,limit_mask=None):
res1 = (self.shear_g1 - model_g1)
res2 = (self.shear_g2 - model_g2)
pl.figure()
pl.subplot(3,1,1)
self.plot_shears(self.shear_g1,self.shear_g2,limit_mask)
pl.subplot(3,1,2)
self.plot_shears(model_g1,model_g2,limit_mask)
pl.subplot(3,1,3)
self.plot_shears(res1 , res2,limit_mask)
def plot_residual_g1g2(self,model_g1,model_g2,limit_mask=None):
res1 = (self.shear_g1 - model_g1)
res2 = (self.shear_g2 - model_g2)
scatter_size = 5
pl.figure()
pl.subplot(3,2,1)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, self.shear_g1 , lw = 0 )
pl.colorbar()
pl.axis('equal')
pl.subplot(3,2,3)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, model_g1 , lw = 0 )
# pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, model_g1 , lw = 0 )
pl.colorbar()
pl.axis('equal')
pl.subplot(3,2,5)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, res1 , lw = 0 )
pl.colorbar()
pl.axis('equal')
pl.subplot(3,2,2)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, self.shear_g2 , lw = 0 )
pl.colorbar()
pl.axis('equal')
pl.subplot(3,2,4)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, model_g2 , lw = 0 )
# pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, model_g2 , lw = 0 )
pl.colorbar()
pl.axis('equal')
pl.subplot(3,2,6)
pl.scatter(self.shear_u_arcmin, self.shear_v_arcmin, scatter_size, res2 , lw = 0 )
pl.colorbar()
pl.axis('equal')
def get_concentr(self,M,z):
# Duffy et al 2008 from King and Mead 2011
concentr = 5.72/(1.+z)**0.71 * (M / 1e14 * cosmology.cospars.h)**(-0.081)
# concentr = 5.72/(1.+z)**0.71 * (M / 1e14)**(-0.081)
return concentr
def draw_model(self,params):
"""
params[0] m200 m_solar
"""
self.n_model_evals +=1
pair_z = np.mean([self.halo_z])
halo_M200 = params[0]
self.nh.M_200= halo_M200
self.nh.update()
self.nh.R_200 = self.nh.r_s*self.nh.concentr
model_g1 , model_g2 , Delta_Sigma , Sigma_Crit, kappa = self.nh.get_shears_with_pz_fast(self.shear_u_arcmin , self.shear_v_arcmin , self.grid_z_centers , self.prob_z, redshift_offset)
limit_mask = np.abs(model_g1 + 1j*model_g2) < weak_limit
return model_g1 , model_g2 , limit_mask , Delta_Sigma , kappa
def log_posterior(self,theta):
model_g1 , model_g2, limit_mask , _ , _ = self.draw_model(theta)
likelihood = self.log_likelihood(model_g1,model_g2,limit_mask)
prior = self.log_prior(theta)
if not np.isfinite(prior):
posterior = -np.inf
else:
# use no info from prior for now
posterior = likelihood
if logger.level == logging.DEBUG:
n_progress = 10
elif logger.level == logging.INFO:
n_progress = 1000
if self.n_model_evals % n_progress == 0:
logger.info('%7d post=% 2.8e like=% 2.8e prior=% 2.4e M200=% 6.3e ' % (self.n_model_evals,posterior,likelihood,prior,theta[0]))
if np.isnan(posterior):
import pdb; pdb.set_trace()
if self.save_all_models:
self.plot_residual_g1g2(model_g1,model_g2,limit_mask)
pl.suptitle('model post=% 10.8e M200=%5.2e' % (posterior,theta[0]) )
filename_fig = 'models/res2.%04d.png' % self.n_model_evals
pl.savefig(filename_fig)
logger.debug('saved %s' % filename_fig)
pl.close()
return posterior
def log_prior(self,theta):
kappa0 = theta[0]
# prob = -0.5 * ( (log10_M200 - self.gaussian_prior_theta[0]['mean'])/self.gaussian_prior_theta[0]['std'] )**2 - np.log(np.sqrt(2*np.pi))
prob=1e-10 # small number so that the prior doesn't matter
if ( self.parameters[0]['box']['min'] <= kappa0 <= self.parameters[0]['box']['max'] ):
return prob
return -np.inf
def log_likelihood(self,model_g1,model_g2,limit_mask):
select = (~np.isnan(model_g1)) & (~np.isnan(model_g2)) & (~np.isinf(model_g1)) & (~np.isinf(model_g2))
res1_sq = ((model_g1[select] - self.shear_g1[select])**2) * self.inv_sq_sigma_g[select] #* limit_mask
res2_sq = ((model_g2[select] - self.shear_g2[select])**2) * self.inv_sq_sigma_g[select] #* limit_mask
# n_points = len(np.nonzero(limit_mask))
# chi2 = -0.5 * ( np.sum( ((res1)/self.sigma_g) **2) + np.sum( ((res2)/self.sigma_g) **2) ) / n_points
chi2 = -0.5 * ( np.sum( res1_sq ) + np.sum( res2_sq ) )
return chi2
def null_log_likelihood(self,h1M200,h2M200):
theta_null = [0,1,h1M200,h2M200]
model_g1 , model_g2, limit_mask , _ , _ = self.draw_model(theta_null)
null_log_like = self.log_likelihood(model_g1,model_g2,limit_mask)
# null_log_post = self.log_posterior(theta_null)
# print null_log_like , null_log_post
# pl.show()
return null_log_like
def run_gridsearch(self):
self.n_model_evals = 0
self.nh = nfw.NfwHalo()
self.nh.z_cluster= self.halo_z
self.nh.theta_cx = self.halo_u_arcmin
self.nh.theta_cy = self.halo_v_arcmin
self.nh.set_mean_inv_sigma_crit(self.grid_z_centers,self.prob_z,self.halo_z)
self.n_model_evals = 0
grid_M200 = np.linspace(self.parameters[0]['box']['min'],self.parameters[0]['box']['max'], self.parameters[0]['n_grid'])
n_total = len(grid_M200)
log_post = np.zeros([len(grid_M200)])
logger.info('running gridsearch total %d grid points' % n_total)
ia = 0
for ik,vk in enumerate(grid_M200):
log_post[ik] = self.log_posterior([grid_M200[ik]])
ia+=1
return log_post , grid_M200
def get_bcc_pz(self,filename_lenscat):
if self.prob_z == None:
# filename_lenscat = os.environ['HOME'] + '/data/BCC/bcc_a1.0b/aardvark_v1.0/lenscats/s2n10cats/aardvarkv1.0_des_lenscat_s2n10.351.fit'
# filename_lenscat = os.environ['HOME'] + '/data/BCC/bcc_a1.0b/aardvark_v1.0/lenscats/s2n10cats/aardvarkv1.0_des_lenscat_s2n10.351.fit'
if 'fits' in filename_lenscat:
lenscat = tabletools.loadTable(filename_lenscat)
if 'z' in lenscat.dtype.names:
self.prob_z , _ = pl.histogram(lenscat['z'],bins=self.grid_z_edges,normed=True)
elif 'z-phot' in lenscat.dtype.names:
self.prob_z , _ = pl.histogram(lenscat['z-phot'],bins=self.grid_z_edges,normed=True)
if 'e1' in lenscat.dtype.names:
select = lenscat['star_flag'] == 0
lenscat = lenscat[select]
select = lenscat['fitclass'] == 0
lenscat = lenscat[select]
select = (lenscat['e1'] != 0.0) * (lenscat['e2'] != 0.0)
lenscat = lenscat[select]
self.sigma_ell = np.std(lenscat['e1']*lenscat['weight'],ddof=1)
elif 'pp2' in filename_lenscat:
pickle = tabletools.loadPickle(filename_lenscat,log=0)
self.prob_z = pickle['prob_z']
self.grid_z_centers = pickle['bins_z']
self.grid_z_edges = plotstools.get_bins_edges(self.grid_z_centers)
def set_shear_sigma(self):
# self.inv_sq_sigma_g = ( np.sqrt(self.shear_n_gals) / self.sigma_ell )**2
self.inv_sq_sigma_g = self.shear_w
# remove nans -- we shouldn't have nans in the data, but we appear to have
select = np.isnan(self.shear_g1) | np.isnan(self.shear_g2)
self.inv_sq_sigma_g[select] = 0
self.shear_g1[select] = 0
self.shear_g2[select] = 0
n_nans = sum(np.isnan(self.shear_g1))
logger.info('found %d nan pixels' % n_nans)