Source code for lisatools.detector

from __future__ import annotations
import os
from abc import ABC, abstractmethod
from typing import Any, List, Tuple, Optional
from dataclasses import dataclass
import requests
from copy import deepcopy
import h5py
from scipy import interpolate

from .utils.constants import *
from .utils.utility import get_array_module

import numpy as np

# import for cpu/gpu
from lisatools.cutils.detector_cpu import pycppDetector as pycppDetector_cpu

try:
    import cupy as cp
    from lisatools.cutils.detector_gpu import pycppDetector as pycppDetector_gpu

except (ImportError, ModuleNotFoundError) as e:
    pycppDetector_gpu = None  # for doc string purposes


SC = [1, 2, 3]
LINKS = [12, 23, 31, 13, 32, 21]

LINEAR_INTERP_TIMESTEP = 600.00  # sec (0.25 hr)


[docs] class Orbits(ABC): """LISA Orbit Base Class Args: filename: File name. File should be in the style of LISAOrbits use_gpu: If ``True``, use a gpu. armlength: Armlength of detector. """ def __init__( self, filename: str, use_gpu: bool = False, armlength: Optional[float] = 2.5e9, ) -> None: self.use_gpu = use_gpu self.filename = filename self.armlength = armlength self._setup() self.configured = False @property def xp(self): """numpy or cupy based on self.use_gpu""" xp = np if not self.use_gpu else cp return xp @property def armlength(self) -> float: """Armlength parameter.""" return self._armlength @armlength.setter def armlength(self, armlength: float) -> None: """armlength setter.""" if isinstance(armlength, float): # TODO: put error check that it is close self._armlength = armlength else: raise ValueError("armlength must be float.") @property def LINKS(self) -> List[int]: """Link order.""" return LINKS @property def SC(self) -> List[int]: """Spacecraft order.""" return SC @property def link_space_craft_r(self) -> List[int]: """Receiver (first) spacecraft""" return [int(str(link_i)[0]) for link_i in self.LINKS] @property def link_space_craft_e(self) -> List[int]: """Sender (second) spacecraft""" return [int(str(link_i)[1]) for link_i in self.LINKS] def _setup(self) -> None: """Read in orbital data from file and store.""" with self.open() as f: for key in f.attrs.keys(): setattr(self, key + "_base", f.attrs[key]) @property def filename(self) -> str: """Orbit file name.""" return self._filename @filename.setter def filename(self, filename: str) -> None: """Set file name.""" assert isinstance(filename, str) if os.path.exists(filename): self._filename = filename else: # get path path_to_this_file = __file__.split("detector.py")[0] # make sure orbit_files directory exists in the right place if not os.path.exists(path_to_this_file + "orbit_files/"): os.mkdir(path_to_this_file + "orbit_files/") path_to_this_file = path_to_this_file + "orbit_files/" if not os.path.exists(path_to_this_file + filename): # download files from github if they are not there github_file = f"https://github.com/mikekatz04/LISAanalysistools/raw/main/lisatools/orbit_files/{filename}" r = requests.get(github_file) # if not success if r.status_code != 200: raise ValueError( f"Cannot find {filename} within default files located at github.com/mikekatz04/LISAanalysistools/lisatools/orbit_files/." ) # write the contents to a local file with open(path_to_this_file + filename, "wb") as f: f.write(r.content) # store self._filename = path_to_this_file + filename
[docs] def open(self) -> h5py.File: """Opens the h5 file in the proper mode. Returns: H5 file object: Opened file. Raises: RuntimeError: If backend is opened for writing when it is read-only. """ f = h5py.File(self.filename, "r") return f
@property def t_base(self) -> np.ndarray: """Time array from file.""" with self.open() as f: t_base = np.arange(self.size_base) * self.dt_base return t_base @property def ltt_base(self) -> np.ndarray: """Light travel times along links from file.""" with self.open() as f: ltt = f["tcb"]["ltt"][:] return ltt @property def n_base(self) -> np.ndarray: """Normal unit vectors towards receiver along links from file.""" with self.open() as f: n = f["tcb"]["n"][:] return n @property def x_base(self) -> np.ndarray: """Spacecraft position from file.""" with self.open() as f: x = f["tcb"]["x"][:] return x @property def v_base(self) -> np.ndarray: """Spacecraft velocities from file.""" with self.open() as f: v = f["tcb"]["v"][:] return v @property def t(self) -> np.ndarray: """Configured time array.""" self._check_configured() return self._t @t.setter def t(self, t: np.ndarray): """Set configured time array.""" assert isinstance(t, np.ndarray) and t.ndim == 1 self._t = t @property def ltt(self) -> np.ndarray: """Light travel time.""" self._check_configured() return self._ltt @ltt.setter def ltt(self, ltt: np.ndarray) -> np.ndarray: """Set light travel time.""" assert ltt.shape[0] == len(self.t) @property def n(self) -> np.ndarray: """Normal vectors along links.""" self._check_configured() return self._n @n.setter def n(self, n: np.ndarray) -> np.ndarray: """Set Normal vectors along links.""" return self._n @property def x(self) -> np.ndarray: """Spacecraft positions.""" self._check_configured() return self._x @x.setter def x(self, x: np.ndarray) -> np.ndarray: """Set Spacecraft positions.""" return self._x @property def v(self) -> np.ndarray: """Spacecraft velocities.""" self._check_configured() return self._v @v.setter def v(self, v: np.ndarray) -> np.ndarray: """Set Spacecraft velocities.""" return self._v
[docs] def configure( self, t_arr: Optional[np.ndarray] = None, dt: Optional[float] = None, linear_interp_setup: Optional[bool] = False, ) -> None: """Configure the orbits to match the signal response generator time basis. The base orbits will be scaled up or down as needed using Cubic Spline interpolation. The higherarchy of consideration to each keyword argument if multiple are given: ``linear_interp_setup``, ``t_arr``, ``dt``. If nothing is provided, the base points are used. Args: t_arr: New time array. dt: New time step. Will take the time duration to be that of the input data. linear_interp_setup: If ``True``, it will create a dense grid designed for linear interpolation with a constant time step. """ x_orig = self.t_base # everything up base on input if linear_interp_setup: # setup spline make_cpp = True dt = LINEAR_INTERP_TIMESTEP Tobs = self.t_base[-1] Nobs = int(Tobs / dt) t_arr = np.arange(Nobs) * dt if t_arr[-1] < self.t_base[-1]: t_arr = np.concatenate([t_arr, self.t_base[-1:]]) elif t_arr is not None: # check array inputs and fill dt assert np.all(t_arr >= self.t_base[0]) and np.all(t_arr <= self.t_base[-1]) make_cpp = True dt = abs(t_arr[1] - t_arr[0]) elif dt is not None: # fill array based on dt and base t make_cpp = True Tobs = self.t_base[-1] Nobs = int(Tobs / dt) t_arr = np.arange(Nobs) * dt if t_arr[-1] < self.t_base[-1]: t_arr = np.concatenate([t_arr, self.t_base[-1:]]) else: make_cpp = False t_arr = self.t_base x_new = t_arr.copy() self.t = t_arr.copy() # use base quantities, and interpolate to prepare new arrays accordingly for which in ["ltt", "x", "n", "v"]: arr = getattr(self, which + "_base") arr_tmp = arr.reshape(self.size_base, -1) arr_out_tmp = np.zeros((len(x_new), arr_tmp.shape[-1])) for i in range(arr_tmp.shape[-1]): arr_out_tmp[:, i] = interpolate.CubicSpline(x_orig, arr_tmp[:, i])( x_new ) arr_out = arr_out_tmp.reshape((len(x_new),) + arr.shape[1:]) setattr(self, "_" + which, arr_out) # make sure base spacecraft and link inormation is ready lsr = np.asarray(self.link_space_craft_r).copy().astype(np.int32) lse = np.asarray(self.link_space_craft_e).copy().astype(np.int32) ll = np.asarray(self.LINKS).copy().astype(np.int32) # indicate this class instance has been configured self.configured = True # prepare cpp class args to load when needed if make_cpp: self.pycppdetector_args = [ dt, len(self.t), self.xp.asarray(self.n.flatten().copy()), self.xp.asarray(self.ltt.flatten().copy()), self.xp.asarray(self.x.flatten().copy()), self.xp.asarray(ll), self.xp.asarray(lsr), self.xp.asarray(lse), self.armlength, ] self.dt = dt else: self.pycppdetector_args = None self.dt = dt
@property def dt(self) -> float: """new time step if it exists""" if self._dt is None: raise ValueError("dt not available for t_arr only.") return self._dt @dt.setter def dt(self, dt: float) -> None: self._dt = dt @property def pycppdetector(self) -> pycppDetector_cpu | pycppDetector_gpu: """C++ class""" if self._pycppdetector_args is None: raise ValueError( "Asking for c++ class. Need to set linear_interp_setup = True when configuring." ) pycppDetector = pycppDetector_cpu if not self.use_gpu else pycppDetector_gpu self._pycppdetector = pycppDetector(*self._pycppdetector_args) return self._pycppdetector @property def pycppdetector_args(self) -> tuple: """args for the c++ class.""" return self._pycppdetector_args @pycppdetector_args.setter def pycppdetector_args(self, pycppdetector_args: tuple) -> None: self._pycppdetector_args = pycppdetector_args @property def size(self) -> int: """Number of time points.""" self._check_configured() return len(self.t) def _check_configured(self) -> None: if not self.configured: raise ValueError( "Cannot request property. Need to use configure() method first." )
[docs] def get_light_travel_times( self, t: float | np.ndarray, link: int | np.ndarray ) -> float | np.ndarray: """Compute light travel time as a function of time. Computes with the c++ backend. Args: t: Time array in seconds. link: which link. Must be ``in self.LINKS``. Returns: Light travel times. """ # test and prepare inputs if isinstance(t, float) and isinstance(link, int): squeeze = True t = self.xp.atleast_1d(t) link = self.xp.atleast_1d(link).astype(np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(link, int): squeeze = False t = self.xp.atleast_1d(t) link = self.xp.full_like(t, link, dtype=np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(link, self.xp.ndarray): squeeze = False t = self.xp.asarray(t) link = self.xp.asarray(link).astype(np.int32) else: raise ValueError( "(t, link) can be (float, int), (np.ndarray, int), (np.ndarray, np.ndarray)." ) # buffer array and c computation ltt_out = self.xp.zeros_like(t) self.pycppdetector.get_light_travel_time_arr_wrap( ltt_out, t, link, len(ltt_out) ) # prepare output if squeeze: return ltt_out[0] return ltt_out
[docs] def get_pos(self, t: float | np.ndarray, sc: int | np.ndarray) -> np.ndarray: """Compute light travel time as a function of time. Computes with the c++ backend. Args: t: Time array in seconds. sc: which spacecraft. Must be ``in self.SC``. Returns: Position of spacecraft. """ # test and setup inputs accordingly if isinstance(t, float) and isinstance(sc, int): squeeze = True t = self.xp.atleast_1d(t) sc = self.xp.atleast_1d(sc).astype(np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(sc, int): squeeze = False t = self.xp.atleast_1d(t) sc = self.xp.full_like(t, sc, dtype=np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(sc, self.xp.ndarray): squeeze = False t = self.xp.asarray(t) sc = self.xp.asarray(sc).astype(np.int32) else: raise ValueError( "(t, sc) can be (float, int), (np.ndarray, int), (np.ndarray, np.ndarray). If the inputs follow this, make sure the orbits class GPU setting matches the arrays coming in (GPU or CPU)." ) # buffer arrays for input into c code pos_x = self.xp.zeros_like(t) pos_y = self.xp.zeros_like(t) pos_z = self.xp.zeros_like(t) # c code computation self.pycppdetector.get_pos_arr_wrap(pos_x, pos_y, pos_z, t, sc, len(pos_x)) # prepare output output = self.xp.array([pos_x, pos_y, pos_z]).T if squeeze: return output.squeeze() return output
[docs] def get_normal_unit_vec( self, t: float | np.ndarray, link: int | np.ndarray ) -> np.ndarray: """Compute link normal vector as a function of time. Computes with the c++ backend. Args: t: Time array in seconds. link: which link. Must be ``in self.LINKS``. Returns: Link normal vectors. """ # test and prepare inputs if isinstance(t, float) and isinstance(link, int): squeeze = True t = self.xp.atleast_1d(t) link = self.xp.atleast_1d(link).astype(np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(link, int): squeeze = False t = self.xp.atleast_1d(t) link = self.xp.full_like(t, link, dtype=np.int32) elif isinstance(t, self.xp.ndarray) and isinstance(link, self.xp.ndarray): squeeze = False t = self.xp.asarray(t) link = self.xp.asarray(link).astype(np.int32) else: raise ValueError( "(t, link) can be (float, int), (np.ndarray, int), (np.ndarray, np.ndarray)." ) # c code with buffers normal_unit_vec_x = self.xp.zeros_like(t) normal_unit_vec_y = self.xp.zeros_like(t) normal_unit_vec_z = self.xp.zeros_like(t) # c code self.pycppdetector.get_normal_unit_vec_arr_wrap( normal_unit_vec_x, normal_unit_vec_y, normal_unit_vec_z, t, link, len(normal_unit_vec_x), ) # prep outputs output = self.xp.array( [normal_unit_vec_x, normal_unit_vec_y, normal_unit_vec_z] ).T if squeeze: return output.squeeze() return output
@property def ptr(self) -> int: """pointer to c++ class""" return self.pycppdetector.ptr
[docs] class EqualArmlengthOrbits(Orbits): """Equal Armlength Orbits Orbit file: equalarmlength-orbits.h5 Args: *args: Arguments for :class:`Orbits`. **kwargs: Kwargs for :class:`Orbits`. """ def __init__(self, *args: Any, **kwargs: Any): super().__init__("equalarmlength-orbits.h5", *args, **kwargs)
[docs] class ESAOrbits(Orbits): """ESA Orbits Orbit file: esa-trailing-orbits.h5 Args: *args: Arguments for :class:`Orbits`. **kwargs: Kwargs for :class:`Orbits`. """ def __init__(self, *args, **kwargs): super().__init__("esa-trailing-orbits.h5", *args, **kwargs)
class DefaultOrbits(EqualArmlengthOrbits): """Set default orbit class to Equal Armlength orbits for now.""" pass
[docs] @dataclass class LISAModelSettings: """Required LISA model settings: Args: Soms_d: OMS displacement noise. Sa_a: Acceleration noise. orbits: Orbital information. name: Name of model. """ Soms_d: float Sa_a: float orbits: Orbits name: str
[docs] class LISAModel(LISAModelSettings, ABC): """Model for the LISA Constellation This includes sensitivity information computed in :py:mod:`lisatools.sensitivity` and orbital information contained in an :class:`Orbits` class object. This class is used to house high-level methods useful to various needed computations. """ def __str__(self) -> str: out = "LISA Constellation Configurations Settings:\n" for key, item in self.__dict__.items(): out += f"{key}: {item}\n" return out def lisanoises( self, f: float | np.ndarray, unit: Optional[str] = "relative_frequency", ) -> Tuple[float, float]: """Calculate both LISA noise terms based on input model. Args: f: Frequency array. unit: Either ``"relative_frequency"`` or ``"displacement"``. Returns: Tuple with acceleration term as first value and oms term as second value. """ # TODO: fix this up Soms_d_in = self.Soms_d Sa_a_in = self.Sa_a frq = f ### Acceleration noise ## In acceleration Sa_a = Sa_a_in * (1.0 + (0.4e-3 / frq) ** 2) * (1.0 + (frq / 8e-3) ** 4) ## In displacement Sa_d = Sa_a * (2.0 * np.pi * frq) ** (-4.0) ## In relative frequency unit Sa_nu = Sa_d * (2.0 * np.pi * frq / C_SI) ** 2 Spm = Sa_nu ### Optical Metrology System ## In displacement Soms_d = Soms_d_in * (1.0 + (2.0e-3 / f) ** 4) ## In relative frequency unit Soms_nu = Soms_d * (2.0 * np.pi * frq / C_SI) ** 2 Sop = Soms_nu if unit == "displacement": return Sa_d, Soms_d elif unit == "relative_frequency": return Spm, Sop
# defaults scirdv1 = LISAModel((15.0e-12) ** 2, (3.0e-15) ** 2, DefaultOrbits(), "scirdv1") proposal = LISAModel((10.0e-12) ** 2, (3.0e-15) ** 2, DefaultOrbits(), "proposal") mrdv1 = LISAModel((10.0e-12) ** 2, (2.4e-15) ** 2, DefaultOrbits(), "mrdv1") sangria = LISAModel((7.9e-12) ** 2, (2.4e-15) ** 2, DefaultOrbits(), "sangria") __stock_list_models__ = [scirdv1, proposal, mrdv1, sangria] __stock_list_models_name__ = [tmp.name for tmp in __stock_list_models__]
[docs] def get_available_default_lisa_models() -> List[LISAModel]: """Get list of default LISA models Returns: List of LISA models. """ return __stock_list_models__
[docs] def get_default_lisa_model_from_str(model: str) -> LISAModel: """Return a LISA model from a ``str`` input. Args: model: Model indicated with a ``str``. Returns: LISA model associated to that ``str``. """ if model not in __stock_list_models_name__: raise ValueError( "Requested string model is not available. See lisatools.detector documentation." ) return globals()[model]
[docs] def check_lisa_model(model: Any) -> LISAModel: """Check input LISA model. Args: model: LISA model to check. Returns: LISA Model checked. Adjusted from ``str`` if ``str`` input. """ if isinstance(model, str): model = get_default_lisa_model_from_str(model) if not isinstance(model, LISAModel): raise ValueError("model argument not given correctly.") return model