Coverage for pesummary/gw/file/formats/lalinference.py: 44.7%
237 statements
« prev ^ index » next coverage.py v7.4.4, created at 2024-12-09 22:34 +0000
1# Licensed under an MIT style license -- see LICENSE.md
3import os
5import h5py
6import numpy as np
8from pesummary.gw.file.formats.base_read import GWRead, GWSingleAnalysisRead
9from pesummary.utils.utils import logger
10from pesummary.utils.decorators import open_config
11from pesummary import conf
13__author__ = ["Charlie Hoy <charlie.hoy@ligo.org>"]
14SAMPLER_KWARGS = {
15 "log_bayes_factor": conf.log_bayes_factor,
16 "bayes_factor": conf.bayes_factor,
17 "log_evidence": conf.log_evidence,
18 "evidence": conf.evidence,
19 "log_prior_volume": conf.log_prior_volume,
20 "sampleRate": "sample_rate",
21 "segmentLength": "segment_length"
22}
24META_DATA = {
25 "flow": "f_low",
26 "f_low": "f_low",
27 "fref": "f_ref",
28 "f_ref": "f_ref",
29 "LAL_PNORDER": "pn_order",
30 "LAL_APPROXIMANT": "approximant",
31 "number_of_live_points": "number_of_live_points",
32 "segmentLength": "segment_length",
33 "segmentStart": "segment_start",
34 "sampleRate": "sample_rate",
35}
38def open_lalinference(path):
39 """Grab the parameters and samples in a lalinference file
41 Parameters
42 ----------
43 path: str
44 path to the result file you wish to read in
45 """
46 from pesummary.core.file.formats.hdf5 import _read_hdf5_with_h5py
47 lalinference_names, samples, _dataset = _read_hdf5_with_h5py(
48 path, return_posterior_dataset=True
49 )
50 if "logdistance" in lalinference_names:
51 lalinference_names.append("luminosity_distance")
52 for num, i in enumerate(samples):
53 samples[num].append(
54 np.exp(i[lalinference_names.index("logdistance")]))
55 if "costheta_jn" in lalinference_names:
56 lalinference_names.append("theta_jn")
57 for num, i in enumerate(samples):
58 samples[num].append(
59 np.arccos(i[lalinference_names.index("costheta_jn")]))
60 extra_kwargs = LALInference.grab_extra_kwargs(path)
61 extra_kwargs["sampler"]["nsamples"] = len(samples)
62 extra_kwargs["sampler"]["pe_algorithm"] = "lalinference"
63 try:
64 version = _dataset.attrs["VERSION"].decode("utf-8")
65 except Exception as e:
66 version = None
67 return {
68 "parameters": lalinference_names,
69 "samples": samples,
70 "injection": None,
71 "version": version,
72 "kwargs": extra_kwargs
73 }
76class LALInference(GWSingleAnalysisRead):
77 """PESummary wrapper of `lalinference`
78 (https://git.ligo.org/lscsoft/lalsuite/lalinference).
80 Parameters
81 ----------
82 path_to_results_file: str
83 path to the results file you wish to load in with `LALInference`
84 remove_nan_likelihood_samples: Bool, optional
85 if True, remove samples which have log_likelihood='nan'. Default True
87 Attributes
88 ----------
89 parameters: list
90 list of parameters stored in the result file
91 converted_parameters: list
92 list of parameters that have been derived from the sampled distributions
93 samples: 2d list
94 list of samples stored in the result file
95 samples_dict: dict
96 dictionary of samples stored in the result file keyed by parameters
97 input_version: str
98 version of the result file passed.
99 extra_kwargs: dict
100 dictionary of kwargs that were extracted from the result file
101 converted_parameters: list
102 list of parameters that have been added
103 pe_algorithm: str
104 name of the algorithm used to generate the posterior samples
106 Methods
107 -------
108 to_dat:
109 save the posterior samples to a .dat file
110 to_latex_table:
111 convert the posterior samples to a latex table
112 generate_latex_macros:
113 generate a set of latex macros for the stored posterior samples
114 to_lalinference:
115 convert the posterior samples to a lalinference result file
116 generate_all_posterior_samples:
117 generate all posterior distributions that may be derived from
118 sampled distributions
119 """
120 def __init__(self, path_to_results_file, injection_file=None, **kwargs):
121 super(LALInference, self).__init__(path_to_results_file, **kwargs)
122 self.load(self._grab_data_from_lalinference_file)
124 @classmethod
125 def load_file(cls, path, injection_file=None, **kwargs):
126 if injection_file and not os.path.isfile(injection_file):
127 raise IOError("%s does not exist" % (path))
128 return super(LALInference, cls).load_file(
129 path, injection_file=injection_file, **kwargs
130 )
132 @staticmethod
133 def guess_path_to_sampler(path):
134 """Guess the path to the sampler group in a LALInference results file
136 Parameters
137 ----------
138 path: str
139 path to the LALInference results file
140 """
141 def _find_name(name):
142 c1 = "lalinference_nest" in name or "lalinference_mcmc" in name
143 c2 = "lalinference_nest/" not in name and "lalinference_mcmc/" not in name
144 if c1 and c2:
145 return name
147 f = h5py.File(path, 'r')
148 _path = f.visit(_find_name)
149 f.close()
150 return _path
152 @staticmethod
153 def _parameters_in_lalinference_file(path):
154 """Return the parameter names stored in the LALInference results file
156 Parameters
157 ----------
158 """
159 f = h5py.File(path, 'r')
160 path_to_samples = GWRead.guess_path_to_samples(path)
161 parameters = list(f[path_to_samples].dtype.names)
162 f.close()
163 return parameters
165 @staticmethod
166 def _samples_in_lalinference_file(path):
167 """
168 """
169 f = h5py.File(path, 'r')
170 path_to_samples = GWRead.guess_path_to_samples(path)
171 samples = [list(i) for i in f[path_to_samples]]
172 return samples
174 @property
175 def calibration_spline_posterior(self):
176 if not any("_spcal_amp" in i for i in self.parameters):
177 return super(LALInference, self).calibration_spline_posterior
178 keys_amp = np.sort(
179 [param for param in self.parameters if "_spcal_amp" in param]
180 )
181 keys_phase = np.sort(
182 [param for param in self.parameters if "_spcal_phase" in param]
183 )
184 IFOs = np.unique(
185 [
186 param.split("_")[0] for param in self.parameters if
187 "_spcal_" in param
188 ]
189 )
190 log_frequencies = {ifo: [] for ifo in IFOs}
191 for key, value in self.extra_kwargs["other"].items():
192 if "_spcal_logfreq" in key:
193 cond = (
194 key.replace("logfreq", "freq") not in
195 self.extra_kwargs["other"].keys()
196 )
197 if cond:
198 log_frequencies[key.split("_")[0]].append(float(value))
199 elif "_spcal_freq" in key:
200 log_frequencies[key.split("_")[0]].append(np.log(float(value)))
201 amp_params = {ifo: [] for ifo in IFOs}
202 phase_params = {ifo: [] for ifo in IFOs}
203 zipped = zip(
204 [keys_amp, keys_phase], [amp_params, phase_params]
205 )
206 _samples = self.samples_dict
207 for keys, dictionary in zipped:
208 for key in keys:
209 ifo = key.split("_")[0]
210 ind = self.parameters.index(key)
211 dictionary[ifo].append(_samples[key])
212 return log_frequencies, amp_params, phase_params
214 @staticmethod
215 def grab_extra_kwargs(path):
216 """Grab any additional information stored in the lalinference file
217 """
218 kwargs = {"sampler": {}, "meta_data": {}, "other": {}}
219 path_to_samples = GWRead.guess_path_to_samples(path)
220 path_to_sampler = LALInference.guess_path_to_sampler(path)
221 f = h5py.File(path, 'r')
223 attributes = dict(f[path_to_sampler].attrs.items())
224 for kwarg, item in attributes.items():
225 if kwarg in list(SAMPLER_KWARGS.keys()) and kwarg == "evidence":
226 kwargs["sampler"][conf.log_evidence] = np.round(np.log(item), 2)
227 elif kwarg in list(SAMPLER_KWARGS.keys()) and kwarg == "bayes_factor":
228 kwargs["sampler"][conf.log_bayes_factor] = np.round(
229 np.log(item), 2
230 )
231 elif kwarg in list(SAMPLER_KWARGS.keys()):
232 kwargs["sampler"][SAMPLER_KWARGS[kwarg]] = np.round(item, 2)
233 else:
234 kwargs["other"][kwarg] = item
236 attributes = dict(f[path_to_samples].attrs.items())
237 for kwarg, item in attributes.items():
238 if kwarg in list(META_DATA.keys()) and kwarg == "LAL_APPROXIMANT":
239 try:
240 from lalsimulation import GetStringFromApproximant
242 kwargs["meta_data"]["approximant"] = \
243 GetStringFromApproximant(
244 int(attributes["LAL_APPROXIMANT"])
245 )
246 except Exception:
247 kwargs["meta_data"]["approximant"] = \
248 int(attributes["LAL_APPROXIMANT"])
249 elif kwarg in list(META_DATA.keys()):
250 kwargs["meta_data"][META_DATA[kwarg]] = item
251 else:
252 kwargs["other"][kwarg] = item
253 f.close()
254 return kwargs
256 @staticmethod
257 def _grab_data_from_lalinference_file(path):
258 """
259 """
260 return open_lalinference(path)
262 def add_fixed_parameters_from_config_file(self, config_file):
263 """Search the conifiguration file and add fixed parameters to the
264 list of parameters and samples
266 Parameters
267 ----------
268 config_file: str
269 path to the configuration file
270 """
271 self._add_fixed_parameters_from_config_file(
272 config_file, self._add_fixed_parameters)
274 def add_marginalized_parameters_from_config_file(self, config_file):
275 """Search the configuration file and add the marginalized parameters
276 to the list of parameters and samples
278 Parameters
279 ----------
280 config_file: str
281 path to the configuration file
282 """
283 self._add_marginalized_parameters_from_config_file(
284 config_file, self._add_marginalized_parameters)
286 @staticmethod
287 @open_config(index=2)
288 def _add_fixed_parameters(parameters, samples, config_file):
289 """Open a LALInference configuration file and add the fixed parameters
290 to the list of parameters and samples
292 Parameters
293 ----------
294 parameters: list
295 list of existing parameters
296 samples: list
297 list of existing samples
298 config_file: str
299 path to the configuration file
300 """
301 from pesummary.gw.file.standard_names import standard_names
303 config = config_file
304 if not config.error:
305 fixed_data = None
306 if "engine" in config.sections():
307 fixed_data = {
308 key.split("fix-")[1]: item for key, item in
309 config.items("engine") if "fix" in key}
310 if fixed_data is not None:
311 for i in fixed_data.keys():
312 fixed_parameter = i
313 fixed_value = fixed_data[i]
314 try:
315 param = standard_names[fixed_parameter]
316 if param in parameters:
317 pass
318 else:
319 parameters.append(param)
320 for num in range(len(samples)):
321 samples[num].append(float(fixed_value))
322 except Exception:
323 if fixed_parameter == "logdistance":
324 if "luminosity_distance" not in parameters:
325 parameters.append(standard_names["distance"])
326 for num in range(len(samples)):
327 samples[num].append(float(fixed_value))
328 if fixed_parameter == "costheta_jn":
329 if "theta_jn" not in parameters:
330 parameters.append(standard_names["theta_jn"])
331 for num in range(len(samples)):
332 samples[num].append(float(fixed_value))
333 return parameters, samples
335 @staticmethod
336 @open_config(index=2)
337 def _add_marginalized_parameters(parameters, samples, config_file):
338 """Open a LALInference configuration file and add the marginalized
339 parameters to the list of parameters and samples
341 Parameters
342 ----------
343 parameters: list
344 list of existing parameters
345 samples: list
346 list of existing samples
347 config_file: str
348 path to the configuration file
349 """
350 config = config_file
351 if not config.error:
352 fixed_data = None
353 if "engine" in config.sections():
354 marg_par = {
355 key.split("marg")[1]: item for key, item in
356 config.items("engine") if "marg" in key}
357 for i in marg_par.keys():
358 if "time" in i and "geocent_time" not in parameters:
359 if "marginalized_geocent_time" in parameters:
360 ind = parameters.index("marginalized_geocent_time")
361 parameters.remove(parameters[ind])
362 parameters.append("geocent_time")
363 for num, j in enumerate(samples):
364 samples[num].append(float(j[ind]))
365 del j[ind]
366 else:
367 logger.warning("You have marginalized over time and "
368 "there are no time samples. Manually "
369 "setting time to 100000s")
370 parameters.append("geocent_time")
371 for num, j in enumerate(samples):
372 samples[num].append(float(100000))
373 if "phi" in i and "phase" not in parameters:
374 if "marginalized_phase" in parameters:
375 ind = parameters.index("marginalized_phase")
376 parameters.remove(parameters[ind])
377 parameters.append("phase")
378 for num, j in enumerate(samples):
379 samples[num].append(float(j[ind]))
380 del j[ind]
381 else:
382 logger.warning("You have marginalized over phase and "
383 "there are no phase samples. Manually "
384 "setting the phase to be 0")
385 parameters.append("phase")
386 for num, j in enumerate(samples):
387 samples[num].append(float(0))
388 if "dist" in i and "luminosity_distance" not in parameters:
389 if "marginalized_distance" in parameters:
390 ind = parameters.index("marginalized_distance")
391 parameters.remove(parameters[ind])
392 parameters.append("luminosity_distance")
393 for num, j in enumerate(samples):
394 samples[num].append(float(j[ind]))
395 del j[ind]
396 else:
397 logger.warning("You have marginalized over distance and "
398 "there are no distance samples. Manually "
399 "setting distance to 100Mpc")
400 parameters.append("luminosity_distance")
401 for num, j in enumerate(samples):
402 samples[num].append(float(100.0))
403 return parameters, samples
404 return parameters, samples
407def _write_lalinference(
408 parameters, samples, outdir="./", label=None, filename=None, overwrite=False,
409 sampler="lalinference_nest", dat=False, **kwargs
410):
411 """Write a set of samples in LALInference file format
413 Parameters
414 ----------
415 parameters: list
416 list of parameters
417 samples: 2d list
418 list of samples. Columns correspond to a given parameter
419 outdir: str
420 The directory where you would like to write the lalinference file
421 label: str
422 The label of the analysis. This is used in the filename if a filename
423 if not specified
424 filename: str
425 The name of the file that you wish to write
426 overwrite: Bool
427 If True, an existing file of the same name will be overwritten
428 sampler: str
429 The sampler which you wish to store in the result file. This may either
430 be 'lalinference_nest' or 'lalinference_mcmc'
431 dat: Bool
432 If True, a LALInference dat file is produced
433 """
434 from pesummary.gw.file.standard_names import lalinference_map
435 from pesummary.utils.samples_dict import SamplesDict
436 import copy
438 _samples = copy.deepcopy(samples)
439 _parameters = copy.deepcopy(parameters)
440 _samples = SamplesDict(_parameters, np.array(_samples).T.tolist())
441 if not filename and not label:
442 from time import time
444 label = round(time())
445 if not filename:
446 extension = "dat" if dat else "hdf5"
447 filename = "lalinference_{}.{}".format(label, extension)
449 if os.path.isfile(os.path.join(outdir, filename)) and not overwrite:
450 raise FileExistsError(
451 "The file '{}' already exists in the directory {}".format(
452 filename, outdir
453 )
454 )
455 reverse_map = {item: key for key, item in lalinference_map.items()}
456 no_key = []
457 for param in _parameters:
458 if param in reverse_map.keys() and reverse_map[param] in _parameters:
459 logger.warning(
460 "The LALInference name for '{}' is '{}'. '{}' already found "
461 "in the posterior table. Keeping both entries".format(
462 param, reverse_map[param], reverse_map[param]
463 )
464 )
465 elif param in reverse_map.keys():
466 _samples[reverse_map[param]] = _samples.pop(param)
467 elif param not in lalinference_map.keys():
468 no_key.append(param)
469 if len(no_key):
470 logger.info(
471 "Unable to find a LALInference name for the parameters: {}. "
472 "Keeping the PESummary name.".format(", ".join(no_key))
473 )
474 lalinference_samples = _samples.to_structured_array()
475 if dat:
476 np.savetxt(
477 os.path.join(outdir, filename), lalinference_samples,
478 delimiter="\t", comments="",
479 header="\t".join(lalinference_samples.dtype.names)
480 )
481 else:
482 with h5py.File(os.path.join(outdir, filename), "w") as f:
483 lalinference = f.create_group("lalinference")
484 sampler = lalinference.create_group(sampler)
485 samples = sampler.create_dataset(
486 "posterior_samples", data=lalinference_samples
487 )
490def write_lalinference(
491 parameters, samples, outdir="./", label=None, filename=None, overwrite=False,
492 sampler="lalinference_nest", dat=False, **kwargs
493):
494 """Write a set of samples in LALInference file format
496 Parameters
497 ----------
498 parameters: list
499 list of parameters
500 samples: 2d list
501 list of samples. Columns correspond to a given parameter
502 outdir: str
503 The directory where you would like to write the lalinference file
504 label: str
505 The label of the analysis. This is used in the filename if a filename
506 if not specified
507 filename: str
508 The name of the file that you wish to write
509 overwrite: Bool
510 If True, an existing file of the same name will be overwritten
511 sampler: str
512 The sampler which you wish to store in the result file. This may either
513 be 'lalinference_nest' or 'lalinference_mcmc'
514 dat: Bool
515 If True, a LALInference dat file is produced
516 """
517 from pesummary.io.write import _multi_analysis_write
519 _multi_analysis_write(
520 _write_lalinference, parameters, samples, outdir=outdir, label=label,
521 filename=filename, overwrite=overwrite, sampler=sampler, dat=dat,
522 file_format="lalinference", **kwargs
523 )