-
Notifications
You must be signed in to change notification settings - Fork 1
/
write_parameter_table.py
258 lines (210 loc) · 12.9 KB
/
write_parameter_table.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
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
.. module:: write_parameter_table
:synopsis: write out a human-readable parameter table file
.. moduleauthor:: Fabian Koehlinger <[email protected]>
Script for writing out a human-readable parameter table file, 'parameter_table.txt', in the base folder
of the specified MontePython chain. This file contains useful info like posterior means, medians and
credibility intervals.
Important: You must have translated your run into a default MontePython chain via
python /path/to/montepython_public/montepython/MontePython.py info /path/to/your/MontePython/chain/{PC, NS, CH}
This script is self-consistent and can be called like this:
python write_parameter_table /path/to/MontePython/chain model_name={'arbitrary string'} sampler={'MH', 'NS', 'MN', 'PC', 'CH'}
various other (mostly plotting-related) options can be set further below!
"""
import os
import sys
import glob
import numpy as np
from utils import minimum_credible_intervals, weighted_mean, quantile
def get_values_and_intervals(parameters, weights, labels, use_median=False):
param_values = np.zeros((len(parameters), 7))
confidence_values = np.zeros((len(parameters), 6))
for idx, param in enumerate(parameters):
print( '-> Calculating histogram for {:}.'.format(labels[idx]))
if use_median:
central_value = quantile(param, [0.5], weights=weights)[0]
else:
central_value = weighted_mean(param, weights=weights)
# bounds returns [[-1sigma, +1sigma],[-2sigma, +2sigma], [-3sigma, +3sigma]]
bounds = minimum_credible_intervals(param, central_value, weights, bins=50)
param_values[idx, :] = np.concatenate(([central_value], bounds[:,0], bounds[:,1]))
confidence_values[idx, :] = central_value + bounds.flatten()
return param_values, confidence_values
def write_best_fit_to_file(fname, best_fit_params, fit_statistics, labels):
"""
Store the bestfit parameters to a file
"""
with open(fname, 'w') as bestfit_file:
bestfit_file.write('# minimal \chi^2 = {:}, index in chain = {:.0f} \n'.format(fit_statistics[0], fit_statistics[3]))
bestfit_file.write(
'# %s\n' % ', '.join(['%16s' % label for label in labels]))
# Removing scale factors in order to store true parameter values
for idx in range(len(labels)):
#bfvalue = chain[a[0], 2+i]*info.scales[i, i]
bf_value = best_fit_params[idx]
if bf_value > 0:
bestfit_file.write(' %.6e\t' % bf_value)
else:
bestfit_file.write('%.6e\t' % bf_value)
bestfit_file.write('\n')
print( 'File saved to: \n', fname)
return
def write_parameters_to_file(fname, best_fit_params, fit_statistics, param_values_mean, confidence_values_mean, param_values_median, confidence_values_median, labels, labels_tex):
with open(fname, 'w') as f:
f.write('# Best fitting values: \n')
f.write('\chi^2 = {:.4f}, \chi^2_red = {:.4f} ({:} d.o.f.), index in chain = {:.0f} \n'.format(fit_statistics[0], fit_statistics[1], int(fit_statistics[2]), fit_statistics[3]))
for index, label in enumerate(labels):
name = label +':'
f.write(name.ljust(20, ' ')+'{:.4f} \n'.format(best_fit_params[index]))
### (weighted) MEAN ###
f.write('\n'+'# parameter, MEAN, err_minus (68%), err_plus (68%), MEAN, err_minus (95%), err_plus (95%), MEAN, err_minus (99%), err_plus (99%) \n')
for index, label in enumerate(labels):
name = label +':'
f.write(name.ljust(20, ' ') + '{0:.4f} {1:.4f} +{2:.4f}, {0:.4f} {3:.4f} +{4:.4f}, {0:.4f} {5:.4f} +{6:.4f} \n'.format(param_values_mean[index, 0], param_values_mean[index, 1], param_values_mean[index, 4], param_values_mean[index, 2], param_values_mean[index, 5], param_values_mean[index, 3], param_values_mean[index, 6]))
f.write('\n'+'# parameter, lower bound (68%), upper bound (68%), lower bound (95%), upper bound (95%), lower bound (99%), upper bound (99%) \n')
for index, label in enumerate(labels):
name = label +':'
f.write(name.ljust(20, ' ')+'1sigma >{:.4f}, 1sigma <{:.4f}, 2sigma >{:.4f}, 2sigma <{:.4f}, 3sigma >{:.4f}, 3sigma <{:.4f} \n'.format(confidence_values_mean[index, 0], confidence_values_mean[index, 1], confidence_values_mean[index, 2], confidence_values_mean[index, 3], confidence_values_mean[index, 4], confidence_values_mean[index, 5]))
### (weighted) MEDIAN ###
f.write('\n'+'# parameter, MEDIAN, err_minus (68%), err_plus (68%), MEDIAN, err_minus (95%), err_plus (95%), MEDIAN, err_minus (99%), err_plus (99%) \n')
for index, label in enumerate(labels):
name = label +':'
f.write(name.ljust(20, ' ') + '{0:.4f} {1:.4f} +{2:.4f}, {0:.4f} {3:.4f} +{4:.4f}, {0:.4f} {5:.4f} +{6:.4f} \n'.format(param_values_median[index, 0], param_values_median[index, 1], param_values_median[index, 4], param_values_median[index, 2], param_values_median[index, 5], param_values_median[index, 3], param_values_median[index, 6]))
f.write('\n'+'# parameter, lower bound (68%), upper bound (68%), lower bound (95%), upper bound (95%), lower bound (99%), upper bound (99%) \n')
for index, label in enumerate(labels):
name = label +':'
f.write(name.ljust(20, ' ')+'1sigma >{:.4f}, 1sigma <{:.4f}, 2sigma >{:.4f}, 2sigma <{:.4f}, 3sigma >{:.4f}, 3sigma <{:.4f} \n'.format(confidence_values_median[index, 0], confidence_values_median[index, 1], confidence_values_median[index, 2], confidence_values_median[index, 3], confidence_values_median[index, 4], confidence_values_median[index, 5]))
### (weighted) MEAN (TeX) ###
f.write('\n'+'\n'+'\n'+'### TeX ###'+'\n'+'# parameter, MEAN, err_minus (68%), err_plus (68%), MEAN, err_minus (95%), err_plus (95%), MEAN, err_minus (99%), err_plus (99%) \n')
for index, label in enumerate(labels_tex):
name = label +':'
f.write(name.ljust(20, ' ')+'{0:.2f}_{{{1:.2f}}}^{{+{2:.2f}}}, {0:.2f}_{{{3:.2f}}}^{{+{4:.2f}}}, {0:.2f}_{{{5:.2f}}}^{{+{6:.2f}}} \n'.format(param_values_mean[index, 0], param_values_mean[index, 1], param_values_mean[index, 4], param_values_mean[index, 2], param_values_mean[index, 5], param_values_mean[index, 3], param_values_mean[index, 6]))
### (weighted) MEDIAN (TeX) ###
f.write('\n'+'\n'+'\n'+'### TeX ###'+'\n'+'# parameter, MEDIAN, err_minus (68%), err_plus (68%), MEDIAN, err_minus (95%), err_plus (95%), MEDIAN, err_minus (99%), err_plus (99%) \n')
for index, label in enumerate(labels_tex):
name = label +':'
f.write(name.ljust(20, ' ')+'{0:.2f}_{{{1:.2f}}}^{{+{2:.2f}}}, {0:.2f}_{{{3:.2f}}}^{{+{4:.2f}}}, {0:.2f}_{{{5:.2f}}}^{{+{6:.2f}}} \n'.format(param_values_median[index, 0], param_values_median[index, 1], param_values_median[index, 4], param_values_median[index, 2], param_values_median[index, 5], param_values_median[index, 3], param_values_median[index, 6]))
print( 'File saved to: \n', fname)
return
def write_table(path_to_chain, model_name='bla', sampler='NS', threshold=0.3):
if sampler == 'NS' or sampler == 'MN':
fnames = [os.path.join(path_to_chain, 'chain_NS__accepted.txt')]
elif sampler == 'MH':
fnames = glob.glob(path_to_chain + '*.txt')
elif sampler == 'CH':
fnames = [os.path.join(path_to_chain, 'chain_CH__sampling.txt')]
elif sampler == 'PC':
fnames = [os.path.join(path_to_chain, 'chain_PC__accepted.txt')]
else:
print( 'You must supply the type of sampler used for the MCMC (MH = Metropolis Hastings, MN = MultiNest, CH = CosmoHammer, PC = PolyChord).')
# deal with multiple chains from MH run and combine them into one (also taking care of burn-in)
counter = 0
for fname in fnames:
if fname not in glob.glob(path_to_chain + '*HEADER.txt') and fname != os.path.join(path_to_chain, 'parameter_table.txt'):
data_tmp = np.loadtxt(fname)
len_chain = data_tmp.shape[0]
idx_gtr_threshold = int(threshold * len_chain)
# remove first 30% of entries as burn-in from MH chain:
# not necessary for NS and CH(?)!
if sampler == 'MH':
data_tmp = data_tmp[idx_gtr_threshold:, :]
if counter == 0:
data = data_tmp
else:
data = np.concatenate((data, data_tmp))
counter += 1
'''
# remove first 30% of entries as burn-in from MH chain:
# not necessary for NS and CH(?)!
if sampler == 'MH':
len_chain = data.shape[0]
idx_gtr_threshold = int(threshold * len_chain)
data = data[idx_gtr_threshold:, :]
'''
weights = data[:, 0]
#print data.shape
#print data[:, -1]
# glob can expand names with *-operator!
fname = glob.glob(path_to_chain + '*_.paramnames')[0]
#print fname
names = np.loadtxt(fname, dtype=str, delimiter='\t')
new_names = names.tolist() #[:-1, :] = names[:, :]
#print np.shape(names)
#print data.shape
added_params = 0
if 'Omega_m ' in names[:, 0] and 'sigma8 ' in names[:, 0]:
idx_Om = np.where('Omega_m ' == names[:, 0])[0]
idx_s8 = np.where('sigma8 ' == names[:, 0])[0]
# +2 because of weights and mloglkl:
S8 = data[:, idx_s8 + 2] * np.sqrt(data[:, idx_Om + 2] / 0.3)
#print S8.mean()
data = np.column_stack((data, S8))
new_names.append(['S8', 'S_{8}'])
added_params += 1
elif 'Omega_m' in names[:, 0] and 'sigma8' in names[:, 0]:
idx_Om = np.where('Omega_m' == names[:, 0])[0]
idx_s8 = np.where('sigma8' == names[:, 0])[0]
# +2 because of weights and mloglkl:
S8 = data[:, idx_s8 + 2] * np.sqrt(data[:, idx_Om + 2] / 0.3)
#print S8.mean()
data = np.column_stack((data, S8))
new_names.append(['S8', 'S_{8}'])
added_params += 1
#exit()
for idx in range(2):
if 'Omega_m_{:} '.format(idx + 1) in names[:, 0] and 'sigma8_{:} '.format(idx + 1) in names[:, 0]:
idx_Om = np.where('Omega_m_{:} '.format(idx + 1) == names[:, 0])[0]
idx_s8 = np.where('sigma8_{:} '.format(idx + 1) == names[:, 0])[0]
# +2 because of weights and mloglkl:
S8 = data[:, idx_s8 + 2] * np.sqrt(data[:, idx_Om + 2] / 0.3)
#print S8.mean()
data = np.column_stack((data, S8))
new_names.append(['S8_{:}'.format(idx + 1), 'S_{{8, \, {:}}}'.format(idx + 1)])
added_params += 1
elif 'Omega_m_{:}'.format(idx + 1) in names[:, 0] and 'sigma8_{:}'.format(idx + 1) in names[:, 0]:
idx_Om = np.where('Omega_m_{:}'.format(idx + 1) == names[:, 0])[0]
idx_s8 = np.where('sigma8_{:}'.format(idx + 1) == names[:, 0])[0]
# +2 because of weights and mloglkl:
S8 = data[:, idx_s8 + 2] * np.sqrt(data[:, idx_Om + 2] / 0.3)
#print S8.mean()
data = np.column_stack((data, S8))
new_names.append(['S8_{:}'.format(idx + 1), 'S_{{8, \, {:}}}'.format(idx + 1)])
added_params += 1
new_names = np.asarray(new_names, dtype=str)
labels = new_names[:, 0]
labels_tex = new_names[:, 1]
for idx, label in enumerate(labels):
if label[-1] == ' ':
labels[idx] = label[:-1]
#print new_names, new_names.shape
#column_names = np.concatenate((np.asarray(['weights', 'mloglkl']), names[:, 0], np.asarray(['S8'])))
chi2 = 2. * data[:, 1]
min_chi2 = chi2.min()
best_fit_index = np.where(data[:, 1] == data[:, 1].min())[0]
#print best_fit_index
#exit()
best_fit_params = data[best_fit_index]
#print data.shape
#print best_fit_params, best_fit_params.shape
#exit()
fit_statistics = np.array([min_chi2, 0., 0., int(best_fit_index[0])])
print( 'Calculating histograms with central value = MEAN.')
params_mean, conf_mean = get_values_and_intervals(data[:, 2:].T, weights, labels, use_median=False)
print( 'Calculating histograms with central value = MEDIAN.')
params_median, conf_median = get_values_and_intervals(data[:, 2:].T, weights, labels, use_median=True)
fname = os.path.join(path_to_chain, 'parameter_table.txt')
write_parameters_to_file(fname, best_fit_params[0, 2:], fit_statistics, params_mean, conf_mean, params_median, conf_median, labels, labels_tex)
fname = os.path.join(path_to_chain, model_name + '.bestfit')
# remove S8 again so that this bestfit file can be used as MP's bestfit-file!!!
write_best_fit_to_file(fname, best_fit_params[0, 2:best_fit_params.size - added_params], fit_statistics, labels[:len(labels) - added_params])
return
if __name__ == '__main__':
path_to_chain = sys.argv[1]
model_name = sys.argv[2]
# needs to be closed with '/' for glob.glob to work properly!
if path_to_chain[-1] != '/':
path_to_chain += '/'
sampler = sys.argv[3]
write_table(path_to_chain, model_name=model_name, sampler=sampler)