read-rapidpe 0.2.0

Read and analyse results generated by rapidpe-rift-pipe

Read Rapid-PE

This is a package to read Rapid-PE outputs.

Install

Install from PyPI

This read-rapidpe package is available on PyPI: https://pypi.org/project/read-rapidpe/

pip install read-rapidpe

Install in dev mode

git clone [e-mail]:yu-kuang.chu/read-rapidpe.git
cd read-rapidpe
pip install -e . 

Example Usage

Reading files

from read_rapidpe import RapidPE_result

run_dir = "path/to/run_dir"
result = RapidPE_result.from_run_dir(run_dir)

There are three optional arguments:

• use_ligolw ( default = True ) : whether to use ligo.lw to read XML files.
• extrinsic_table ( default = True ) : whether to load extrinsic parameter as well.
• parallel_n ( default = 1 ) : number of parallel jobs when reading XML files.

For example, one can do the following to speed up the reading process:

result = RapidPE_result.from_run_dir(run_dir, use_ligolw=False, extrinsic_table=False, parallel_n=4)

Plot marginalized log-likelihood on m1-m2 grid points

import matplotlib.pyplot as plt

# Plot marginalized-log-likelihood over intrinsic parameter (mass_1/mass_2) grid points
plt.scatter(result.mass_1, result.mass_2, c=result.marg_log_likelihood )
plt.xlabel("$m_1$")
plt.ylabel("$m_2$")
plt.colorbar(label="$\ln(L_{marg})$")

Plot interpolated likelihood

import matplotlib.pyplot as plt
import numpy as np


# Create Random m1, m2 samples
m1 = np.random.random(10000)*5
m2 = np.random.random(10000)*5


# After calling result.do_interpolate_marg_log_likelihood_m1m2(), 
# the method result.log_likelihood(m1, m2) will be avalible.
result.do_interpolate_marg_log_likelihood_m1m2()

# Calculate interpolated log_likelihood
log_likelihood = result.log_likelihood(m1, m2)


# =============== Plotting ===============
# Plot interpolated likelihood 
plt.scatter(m1, m2, c=np.exp(log_likelihood), marker=".", s=3, alpha=0.1)

# Plot marginalized likelihood on grid points
plt.scatter(result.mass_1, result.mass_2, c=np.exp(result.marg_log_likelihood), marker="+", vmin=0)

plt.xlabel("$m_1$")
plt.ylabel("$m_2$")
plt.colorbar(label=r"$\mathcal{L}$")

Convert to Pandas DataFrame

import pandas as pd
from read_rapidpe import RapidPE_grid_point


grid_point = RapidPE_grid_point.from_xml("ILE_iteration_xxxxxxxxxx.samples.xml.gz")
pd.DataFrame(grid_point.extrinsic_table)