Skip to content

Commit

Permalink
use suffix to simplify zeroecc handling
Browse files Browse the repository at this point in the history
  • Loading branch information
@md-arif-shaikh
md-arif-shaikh committed Nov 25, 2024
1 parent f5c32c1 commit 41dfcb8
Show file tree
Hide file tree
Showing 3 changed files with 112 additions and 94 deletions.
180 changes: 94 additions & 86 deletions gw_eccentricity/eccDefinition.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@

import numpy as np
import matplotlib.pyplot as plt
import warnings
from .load_data import get_coprecessing_data_dict
from .utils import peak_time_via_quadratic_fit, check_kwargs_and_set_defaults
from .utils import amplitude_using_all_modes
Expand Down Expand Up @@ -169,10 +170,11 @@ def __init__(self, dataDict, num_orbits_to_exclude_before_merger=2,
directly via `dataDict`.
- Set `frame="inertial"` and provide the inertial frame modes
via `dataDict`. In this case, the modes in `dataDict` are
rotated internally (see `rotate_modes`) before further
computation. To get the coprecessing modes from the inertial
modes accurately, `dataDict` must include all modes for
`ell=2`, i.e., (2, -2), (2, -1), (2, 0), (2, 1) and (2, 2).
rotated internally (see `transform_inertial_to_coprecessing`)
before further computation. To get the coprecessing modes
from the inertial modes accurately, `dataDict` must include
all modes for `ell=2`, i.e., (2, -2), (2, -1), (2, 0), (2, 1)
and (2, 2).
For nonprecessing systems, the inertial and coprecessing frames are
equivalent, so there is no distinction. For nonprecessing systems,
Expand Down Expand Up @@ -621,7 +623,18 @@ def process_data_dict(self,
# frame, rotate the modes and obtain the corresponding modes in the
# coprecessing frame
if self.precessing is True and self.frame == "inertial":
dataDict = self.rotate_modes(dataDict)
warnings.warn(
f"The system is precessing but the modes are provided in "
f"the {self.frame} frame. Transforming the modes from"
f" the {self.frame} frame to the coprecessing frame and "
"updating `self.frame` to `coprecessing`.")
dataDict = self.transform_inertial_to_coprecessing(dataDict)
# transform the zeroecc modes as well if provided in dataDict
if "hlm_zeroecc" in dataDict or "amplm_zeroecc" in dataDict:
dataDict = self.transform_inertial_to_coprecessing(
dataDict, suffix="_zeroecc")
# Now that the modes are in the coprecessing frame, update frame
self.frame = "coprecessing"
# Create a new dictionary that will contain the data necessary for
# eccentricity measurement.
newDataDict = {}
Expand Down Expand Up @@ -701,93 +714,88 @@ def process_data_dict(self,
:index_num_orbits_earlier_than_merger]
return newDataDict, t_merger, amp_gw_merger, min_width_for_extrema

def rotate_modes(self, data_dict):
""""Rotate intertial frame modes to obtain coprecessing frame modes.
def transform_inertial_to_coprecessing(self, data_dict, suffix=""):
""""Transfrom intertial frame modes to coprecessing frame modes.
Parameters
----------
data_dict: dict
Dictionary of modes in the inertial frame. The modes are rotated
using `get_coprecessing_data_dict` to obtain the corresponding
modes in the coprecessing frame.
To obtain coprecessing frame modes from an inertial frame dictionary,
this dictionary should include all (ell, m) modes for a specified
ell value. For instance, the inertial modes dictionary should
contain at least all the modes for `ell=2`, i.e., (2, -2), (2, -1),
(2, 0), (2, 1), and (2, 2), to achieve accurate coprecessing frame
modes.
Dictionary of modes in the inertial frame. The modes are
transformed using `get_coprecessing_data_dict` to obtain the
corresponding modes in the coprecessing frame.
To obtain coprecessing frame modes from an inertial frame
dictionary, this dictionary should include all (ell, m) modes for a
specified ell value. For instance, the inertial modes dictionary
should contain at least all the modes for `ell=2`, i.e., (2, -2),
(2, -1), (2, 0), (2, 1), and (2, 2), to achieve accurate
coprecessing frame modes.
suffix: str, default=""
A suffix used to specify which input modes dictionary to use for
obtaining the coprecessing modes. For example, using
`suffix="_zeroecc"` selects the input modes corresponding to the
"zeroecc" modes dictionary, which are then rotated to compute the
coprecessing modes. If left as the default value (`""`), the input
modes from the eccentric modes dictionary are used.
Returns
-------
data_dict with the inertial modes replaced by the corresponding
modes in the coprecessing frame.
"""
if "hlm" in data_dict:
data_dict = get_coprecessing_data_dict(data_dict)
if "hlm_zeroecc" in data_dict:
# `get_coprecessing_data_dict` looks for modes using the key "hlm"
# and the associated time using "t".
# Make a dictionary with a key "hlm" and pass the zeroecc modes
# dict via it. Similary for the times using "t".
intertial_zeroecc_data_dict = {"t": data_dict["t_zeroecc"],
"hlm": data_dict["hlm_zeroecc"]}
coprecessing_zeroecc_data_dict = get_coprecessing_data_dict(
intertial_zeroecc_data_dict)
# update the data_dict by replacing the inertial frame zeroecc
# modes with coprecessing frame zeroecc modes.
data_dict.update(
{"hlm_zeroecc": coprecessing_zeroecc_data_dict["hlm"]})

if ("hlm" + suffix) in data_dict:
data_dict = get_coprecessing_data_dict(data_dict, suffix=suffix)
# if hlm is not in data_dict, get it from amplm and phaselm.
# We need to provide hlm to get the rotated modes.
if "hlm" not in data_dict:
if ("hlm" + suffix) not in data_dict:
amplm_dict = self.get_amplm_from_dataDict(data_dict)
phaselm_dict = self.get_phaselm_from_dataDict(data_dict)
hlm_dict = {}
# check if "hlm_zeroecc" is not in data_dict but amplm_zeroecc is
# in data_dict
if "amplm_zeroecc" in data_dict and "hlm_zeroecc" not in data_dict:
add_hlm_zeroecc = True
hlm_zeroecc_dict = {}
else:
add_hlm_zeroecc = False
# combine amplm and phaselm to get hlm
for k in amplm_dict["amplm"]:
hlm_dict.update({k: amplm_dict["amplm"][k]
* np.exp(-1j * phaselm_dict["phaselm"][k])})
if add_hlm_zeroecc:
hlm_zeroecc_dict.update(
{k: amplm_dict["amplm_zeroecc"][k]
* np.exp(-1j * phaselm_dict["phaselm_zeroecc"][k])})
inertial_ecc_data_dict = {"t": data_dict["t"], "hlm": hlm_dict}
coprecessing_ecc_data_dict = get_coprecessing_data_dict(
inertial_ecc_data_dict)
data_dict.update(coprecessing_ecc_data_dict)
# remove amplm, phaselm because these are in the inertial frame
# and are given priority over hlm when using data for egw
data_dict.pop("amplm", None)
data_dict.pop("phaselm", None)
if add_hlm_zeroecc:
inertial_zeroecc_data_dict = {
"t": data_dict["t_zeroecc"], "hlm": hlm_zeroecc_dict}
coprecessing_zeroecc_data_dict = get_coprecessing_data_dict(
inertial_zeroecc_data_dict)
data_dict.update(
{"hlm_zeroecc": coprecessing_zeroecc_data_dict["hlm"]})
# remove amplm_zeroecc, phaselm_zeroecc because these are in
# the inertial frame and are given priority over hlm_zeroecc
# when using data for egw
data_dict.pop("amplm_zeroecc", None)
data_dict.pop("phaselm_zeroecc", None)
# remove inertial omegalm. The coprecessing omegalm should be computed
# from the coprecessing phase later.
for omega in ["omegalm", "omegalm_zeroecc"]:
if omega in data_dict:
data_dict.pop(omega, None)
hlm_dict = self.get_hlm_from_amplm_phaselm(amplm_dict, phaselm_dict)
data_dict.update(hlm_dict)
data_dict = get_coprecessing_data_dict(data_dict, suffix=suffix)
warnings.warn(
f"Removing the inertial {'amplm' + suffix}, "
f"{'phaselm' + suffix} from `dataDict`. The same is computed "
f"later from the coprecessing {'hlm' + suffix} in"
f"`get_amp_phase_omega_data.`")
data_dict.pop("amplm" + suffix, None)
data_dict.pop("phaselm" + suffix, None)
if "omegalm" + suffix in data_dict:
warnings.warn(
f"Removing the inertial {'omegalm' + suffix} from `dataDict`. "
f"The coprecessing {'omegalm' + suffix}. The same is computed "
f"later from the coprecessing {'hlm' + suffix} in "
f"`get_amp_phase_omega_data`.")
data_dict.pop("omegalm" + suffix, None)
return data_dict

def get_hlm_from_amplm_phaselm(self, amplm_dict, phaselm_dict):
"""Compute the complex hlm modes from amplitude and phase data.
The hlm modes are calculated using the formula: hlm = amplm * exp(-1j *
phaselm), where `amplm` and `phaselm` are dictionaries containing the
amplitude and phase for each mode.
If the `amplm` and `phaselm` dictionaries include "zeroecc" modes,
these are also processed, and the corresponding hlm_zeroecc modes are
added to the returned `data_dict`.
"""
if "amplm_zeroecc" in amplm_dict:
suffixes = ["", "_zeroecc"]
else:
suffixes = [""]
data_dict = {}
for suffix in suffixes:
data_dict["hlm" + suffix] = {}
for k in amplm_dict["amplm" + suffix]:
data_dict["hlm" + suffix].update(
{k: amplm_dict["amplm" + suffix][k]
* np.exp(-1j * phaselm_dict["phaselm" + suffix][k])})
return data_dict

def get_amp_phase_omega_gw(self, data_dict, data_name_suffix=""):
def get_amp_phase_omega_gw(self, data_dict, suffix=""):
"""Get the gw quanitities from modes dict in the coprecessing frame.
For nonprecessing systems, the amp_gw, phase_gw and omega_gw are the same
Expand Down Expand Up @@ -816,26 +824,26 @@ def get_amp_phase_omega_gw(self, data_dict, data_name_suffix=""):
A dictionary with the waveform modes and times. The structure of
`data_dict` should be consistent with that of `dataDict`.
data_name_suffix: str, default=""
suffix: str, default=""
A string identifier for selecting the data type when accessing
amplitude, phase, and frequency (omega) values. It is used to
specify if the waveform data corresponds to eccentric modes or
zero-eccentricity (zeroecc) modes. For instance, set
`data_name_suffix=""` for eccentric modes or
`data_name_suffix="zeroecc"` for non-eccentric modes.
`suffix=""` for eccentric modes or
`suffix="_zeroecc"` for non-eccentric modes.
It will look for the key=`data_name` if `data_name_suffix` is empty
else key=`data_name` + "_" + "data_name_suffix" where the
It will look for the key=`data_name` if `suffix` is empty
else key=`data_name` + `suffix` where the
`data_name` are "amplm", "phaselm" or "omegalm".
"""
# get the correct keys
if data_name_suffix == "":
if suffix == "":
t_key, amp_key, phase_key, omega_key = "t", "amplm", "phaselm", "omegalm"
else:
t_key = "t" + "_" + data_name_suffix
amp_key = "amplm" + "_" + data_name_suffix
phase_key = "phaselm" + "_" + data_name_suffix
omega_key = "omegalm" + "_" + data_name_suffix
t_key = "t" + suffix
amp_key = "amplm" + suffix
phase_key = "phaselm" + suffix
omega_key = "omegalm" + suffix
if not self.precessing:
amp_gw, phase_gw, omega_gw = (data_dict[amp_key][(2, 2)],
data_dict[phase_key][(2, 2)],
Expand Down Expand Up @@ -2170,7 +2178,7 @@ def compute_res_amp_gw_and_res_omega_gw(self):
# get amplitude, phase and omega omega data
self.amp_gw_zeroecc, self.phase_gw_zeroecc, self.omega_gw_zeroecc\
= self.get_amp_phase_omega_gw(self.dataDict,
data_name_suffix="zeroecc")
suffix="_zeroecc")
# to get the residual amplitude and omega, we need to shift the
# zeroecc time axis such that the merger of the zeroecc is at the
# same time as that of the eccentric waveform
Expand Down
9 changes: 5 additions & 4 deletions gw_eccentricity/gw_eccentricity.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -311,10 +311,11 @@ def measure_eccentricity(tref_in=None,
directly via `dataDict`.
- Set `frame="inertial"` and provide the inertial frame modes
via `dataDict`. In this case, the modes in `dataDict` are
rotated internally (see `rotate_modes`) before further
computation. To get the coprecessing modes from the inertial
modes accurately, `dataDict` must include all modes for
`ell=2`, i.e., (2, -2), (2, -1), (2, 0), (2, 1) and (2, 2).
rotated internally (see `transform_inertial_to_coprecessing`)
before further computation. To get the coprecessing modes from
the inertial modes accurately, `dataDict` must include all
modes for `ell=2`, i.e., (2, -2), (2, -1), (2, 0), (2, 1) and
(2, 2).
For nonprecessing systems, the inertial and coprecessing frames are
equivalent, so there is no distinction. For nonprecessing systems, it
Expand Down
17 changes: 13 additions & 4 deletions gw_eccentricity/load_data.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1534,7 +1534,7 @@ def unpack_scri_modes(w):
return result


def get_coprecessing_data_dict(data_dict, ell_min=2, ell_max=2):
def get_coprecessing_data_dict(data_dict, ell_min=2, ell_max=2, suffix=""):
"""Get `data_dict` in the coprecessing frame.
Given a `data_dict` containing the modes dict in the inertial frame and the
Expand All @@ -1556,6 +1556,14 @@ def get_coprecessing_data_dict(data_dict, ell_min=2, ell_max=2):
ell_max: int, default=2
Maximum `ell` value to use.
suffix: str, default=""
A suffix used to specify which input modes dictionary to use for
obtaining the coprecessing modes. For example, using
`suffix="_zeroecc"` selects the input modes corresponding to the
"zeroecc" modes dictionary, which are then rotated to compute the
coprecessing modes. If left as the default value (`""`), the input
modes from the eccentric modes dictionary are used.
Returns
-------
Dictionary of waveform modes in the coprecessing frame and the associated
Expand All @@ -1564,13 +1572,13 @@ def get_coprecessing_data_dict(data_dict, ell_min=2, ell_max=2):
"""
# Get list of modes from `data_dict` to use as input to `scri.WaveformModes`.
ordered_mode_list = package_modes_for_scri(
data_dict["hlm"],
data_dict["hlm" + suffix],
ell_min=ell_min,
ell_max=ell_max)

w = scri.WaveformModes(
dataType=scri.h,
t=data_dict["t"],
t=data_dict["t" + suffix],
data=ordered_mode_list,
ell_min=ell_min,
ell_max=ell_max,
Expand All @@ -1583,7 +1591,8 @@ def get_coprecessing_data_dict(data_dict, ell_min=2, ell_max=2):
# Create a copy of data_dict and replace the "hlm" modes in the inertial frame
# with the corresponding modes in the coprecessing frame
data_dict_copr = deepcopy(data_dict)
data_dict_copr.update({"hlm": unpack_scri_modes(deepcopy(w_coprecessing))})
data_dict_copr.update(
{"hlm" + suffix: unpack_scri_modes(deepcopy(w_coprecessing))})
return data_dict_copr


Expand Down

0 comments on commit 41dfcb8

Please sign in to comment.