def to_file(self, outdir, label):
""" Write the prior distribution to file.
Parameters
----------
outdir: str
output directory name
label: str
Output file naming scheme
"""
check_directory_exists_and_if_not_mkdir(outdir)
prior_file = os.path.join(outdir, "{}.prior".format(label))
logger.debug("Writing priors to {}".format(prior_file))
joint_dists = []
with open(prior_file, "w") as outfile:
for key in self.keys():
if JointPrior in self[key].__class__.__mro__:
distname = '_'.join(self[key].dist.names) + '_{}'.format(self[key].dist.distname)
if distname not in joint_dists:
joint_dists.append(distname)
outfile.write(
"{} = {}\n".format(distname, self[key].dist))
diststr = repr(self[key].dist)
priorstr = repr(self[key])
outfile.write(
"{} = {}\n".format(key, priorstr.replace(diststr,
distname)))
else:
outfile.write(
"{} = {}\n".format(key, self[key]))
def to_json(self, outdir, label):
check_directory_exists_and_if_not_mkdir(outdir)
prior_file = os.path.join(outdir, "{}_prior.json".format(label))
logger.debug("Writing priors to {}".format(prior_file))
with open(prior_file, "w") as outfile:
json.dump(self._get_json_dict(), outfile, cls=BilbyJsonEncoder,
indent=2)
def _parse_and_load():
args, unknown_args = parse_args(get_command_line_arguments(),
create_parser())
logger.info(f"Generating plots for results file {args.result}")
result = CBCResult.from_json(args.result)
if "data_dump" in result.meta_data and os.path.exists(
result.meta_data["data_dump"]):
data_dump = DataDump.from_pickle(result.meta_data["data_dump"])
else:
data_dump = None
if hasattr(args, "webdir"):
outdir = os.path.join(args.webdir, "bilby")
elif hasattr(args, "outdir"):
outdir = args.outdir
else:
outdir = result.outdir
logger.info(f"Plots will be made in {outdir}")
check_directory_exists_and_if_not_mkdir(outdir)
result.outdir = outdir
return args, result, data_dump
#!/usr/bin/env python
"""
An example of how to use bilby to perform parameter estimation for hyper params
"""
import matplotlib.pyplot as plt
import numpy as np
from bilby.core.likelihood import GaussianLikelihood
from bilby.core.prior import Uniform
from bilby.core.result import make_pp_plot
from bilby.core.sampler import run_sampler
from bilby.core.utils import check_directory_exists_and_if_not_mkdir
from bilby.hyper.likelihood import HyperparameterLikelihood
outdir = "outdir"
check_directory_exists_and_if_not_mkdir(outdir)
# Define a model to fit to each data set
def model(x, c0, c1):
return c0 + c1 * x
N = 10
x = np.linspace(0, 10, N)
sigma = 1
Nevents = 4
labels = ["a", "b", "c", "d"]
true_mu_c0 = 5
true_sigma_c0 = 1