Source code for synthesizer.load_data.load_eagle

"""
A module for interfacing with the outputs of the EAGLE
hydrodynamical simulations suite.

The EAGLE hdf5 data loading scripts have been taken from
the `eagle_io` package (https://github.com/flaresimulations/eagle_IO/).
This has been to make the call as self-contained as possible.

The functions here used in
https://github.com/flaresimulations/synthesizer-pipeline

"""

import glob
import re
import timeit
from collections import namedtuple
from functools import partial
from typing import Any, Dict, List, Union

import h5py
import numpy as np
from astropy.cosmology import LambdaCDM
from numpy.typing import NDArray
from typing_extensions import Never

from synthesizer.exceptions import InconsistentArguments, UnmetDependency

try:
    import schwimmbad
except ImportError:
    raise UnmetDependency(
        "The `schwimmbad` module is required to load eagle data. "
        "Install synthesizer with the `eagle` extra dependencies:"
        " `pip install .[eagle]`."
    )

from unyt import Mpc, Msun, unyt_array, unyt_quantity, yr

from synthesizer.load_data.utils import lookup_age

from ..particle.galaxy import Galaxy

# define EAGLE cosmology
cosmo = LambdaCDM(Om0=0.307, Ode0=0.693, H0=67.77, Ob0=0.04825)
norm = np.linalg.norm


[docs] def load_EAGLE( fileloc: str, args: namedtuple, tot_chunks: int = 1535, verbose: bool = False, ) -> List[Union[Galaxy, Never]]: """ Load EAGLE required EAGLE galaxy properties for generating their SEDs Most useful for running on high-z snaps or individual chunks Args: fileloc (string): eagle data file location args (namedtuple): parser arguments passed on to this job tot_chunks (int) total number of files to process verbose (bool) Are we talking? Returns: a dictionary of Galaxy objects with stars and gas components """ tag = args.tag numThreads = args.nthreads chunk = args.chunk if chunk > tot_chunks: raise InconsistentArguments( "Value specified by 'chunk' should be lower" "than the total chunks (tot_chunks)" ) # Read some of the simulation box information h = read_header("SUBFIND", fileloc, tag, "HubbleParam") boxl = read_header("SUBFIND", fileloc, tag, "BoxSize") / h exp_fac = read_header("SUBFIND", fileloc, tag, "ExpansionFactor") zed = (1.0 / exp_fac) - 1.0 boxl = boxl / (1 + zed) with h5py.File( f"{fileloc}/groups_{tag}/eagle_subfind_tab_{tag}.{chunk}.hdf5", "r" ) as hf: sgrpno = np.array(hf.get("/Subhalo/SubGroupNumber"), dtype=np.int32) grpno = np.array(hf.get("/Subhalo/GroupNumber"), dtype=np.int32) # centre of potential of subhalo in h-less physical Mpc cop = np.array( hf.get("/Subhalo/CentreOfPotential"), dtype=np.float32 ) / (h * (1 + zed)) if grpno.dtype == object: return [] # Get required star particle properties s_sgrpno = read_array( "PARTDATA", fileloc, tag, "/PartType4/SubGroupNumber", numThreads=numThreads, verbose=verbose, ) s_grpno = read_array( "PARTDATA", fileloc, tag, "/PartType4/GroupNumber", numThreads=numThreads, ) # Create mask for particles in the current chunk ok = np.in1d(s_sgrpno, sgrpno) * np.in1d(s_grpno, grpno) s_sgrpno = s_sgrpno[ok] s_grpno = s_grpno[ok] s_coords = read_array( "PARTDATA", fileloc, tag, "/PartType4/Coordinates", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] # physical Mpc s_imasses = ( read_array( "PARTDATA", fileloc, tag, "/PartType4/InitialMass", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] * 1e10 ) # Msun s_masses = ( read_array( "PARTDATA", fileloc, tag, "/PartType4/Mass", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] * 1e10 ) # Msun s_ages = read_array( "PARTDATA", fileloc, tag, "/PartType4/StellarFormationTime", numThreads=numThreads, verbose=verbose, )[ok] s_ages = get_age(s_ages, zed, args, numThreads=numThreads) # Gyr s_Zsmooth = read_array( "PARTDATA", fileloc, tag, "/PartType4/SmoothedMetallicity", numThreads=numThreads, verbose=verbose, )[ok] s_hsml = read_array( "PARTDATA", fileloc, tag, "/PartType4/SmoothingLength", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] # physical Mpc s_oxygen = read_array( "PARTDATA", fileloc, tag, "/PartType4/ElementAbundance/Oxygen", numThreads=numThreads, verbose=verbose, )[ok] s_hydrogen = read_array( "PARTDATA", fileloc, tag, "/PartType4/ElementAbundance/Hydrogen", numThreads=numThreads, verbose=verbose, )[ok] # Get gas particle properties g_sgrpno = read_array( "PARTDATA", fileloc, tag, "/PartType0/SubGroupNumber", numThreads=numThreads, verbose=verbose, ) g_grpno = read_array( "PARTDATA", fileloc, tag, "/PartType0/GroupNumber", numThreads=numThreads, verbose=verbose, ) # Create mask for particles in the current chunk ok = np.in1d(g_sgrpno, sgrpno) * np.in1d(g_grpno, grpno) g_sgrpno = g_sgrpno[ok] g_grpno = g_grpno[ok] g_coords = read_array( "PARTDATA", fileloc, tag, "/PartType0/Coordinates", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] # physical Mpc g_masses = ( read_array( "PARTDATA", fileloc, tag, "/PartType0/Mass", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] * 1e10 ) # Msun g_sfr = read_array( "PARTDATA", fileloc, tag, "/PartType0/StarFormationRate", numThreads=numThreads, verbose=verbose, )[ok] # Msol / yr g_Zsmooth = read_array( "PARTDATA", fileloc, tag, "/PartType0/SmoothedMetallicity", numThreads=numThreads, verbose=verbose, )[ok] g_hsml = read_array( "PARTDATA", fileloc, tag, "/PartType0/SmoothingLength", noH=True, physicalUnits=True, numThreads=numThreads, verbose=verbose, )[ok] # physical Mpc _f = partial( assign_galaxy_prop, zed=zed, boxl=boxl, aperture=args.aperture, grpno=grpno, sgrpno=sgrpno, cop=cop, s_grpno=s_grpno, s_sgrpno=s_sgrpno, s_imasses=s_imasses, s_masses=s_masses, s_ages=s_ages, s_Zsmooth=s_Zsmooth, s_coords=s_coords, s_hsml=s_hsml, s_oxygen=s_oxygen, s_hydrogen=s_hydrogen, g_grpno=g_grpno, g_sgrpno=g_sgrpno, g_masses=g_masses, g_Zsmooth=g_Zsmooth, g_sfr=g_sfr, g_coords=g_coords, g_hsml=g_hsml, verbose=verbose, ) if numThreads == 1: pool = schwimmbad.SerialPool() elif numThreads == -1: pool = schwimmbad.MultiPool() else: pool = schwimmbad.MultiPool(processes=numThreads) with schwimmbad.MultiPool(numThreads) as pool: galaxies = pool.map(_f, np.arange(len(grpno))) return galaxies
[docs] def load_EAGLE_shm( chunk: int, fileloc: str, tag: str, s_len: int, g_len: int, args: namedtuple, numThreads: int = 1, dtype: str = "float32", tot_chunks: int = 1535, verbose: bool = False, ) -> List[Union[Galaxy, Never]]: """ Load EAGLE required EAGLE galaxy properties for generating their SEDs using numpy memmap. Most useful for running on high-z snaps Args: chunk (int) file number to process fileloc (string): eagle data file location tag (string): snapshot tag to load s_len (int): total number of star particles g_len (int): total number of gas particles args (namedtuple): parser arguments passed on to this job numThreads (int) number of threads to use dtype (numpy object): data type of the array in memory tot_chunks (int) total number of files to process verbose (bool) Are we talking? Returns: a dictionary of Galaxy objects with stars and gas components """ # get the redshift from the given eagle tag zed = float(tag[5:].replace("p", ".")) h = 0.6777 if chunk > tot_chunks: InconsistentArguments( "Value specified by 'chunk' should be lower" "than the total chunks (tot_chunks)" ) # Read some of the simulation box information h = read_header("SUBFIND", fileloc, tag, "HubbleParam") boxl = read_header("SUBFIND", fileloc, tag, "BoxSize") / h exp_fac = read_header("SUBFIND", fileloc, tag, "ExpansionFactor") zed = (1.0 / exp_fac) - 1.0 with h5py.File( f"{fileloc}/groups_{tag}/eagle_subfind_tab_{tag}.{chunk}.hdf5", "r" ) as hf: sgrpno = np.array(hf.get("/Subhalo/SubGroupNumber"), dtype=np.int32) grpno = np.array(hf.get("/Subhalo/GroupNumber"), dtype=np.int32) # centre of potential of subhalo in h-less physical Mpc cop = np.array( hf.get("/Subhalo/CentreOfPotential"), dtype=np.float32 ) / (h * (1 + zed)) if grpno.dtype == object: return [] # read in required star particle shared memory arrays s_grpno = np_shm_read( f"{args.shm_prefix}s_grpno{args.shm_suffix}", (s_len,), "int32" ) s_sgrpno = np_shm_read( f"{args.shm_prefix}s_sgrpno{args.shm_suffix}", (s_len,), "int32" ) # Create mask for particles in the current chunk ok = np.in1d(s_sgrpno, sgrpno) * np.in1d(s_grpno, grpno) s_sgrpno = s_sgrpno[ok] s_grpno = s_grpno[ok] s_imasses = np_shm_read( f"{args.shm_prefix}s_imasses{args.shm_suffix}", (s_len,), dtype )[ok] s_masses = np_shm_read( f"{args.shm_prefix}s_masses{args.shm_suffix}", (s_len,), dtype )[ok] s_ages = np_shm_read( f"{args.shm_prefix}s_ages{args.shm_suffix}", (s_len,), dtype )[ok] s_Zsmooth = np_shm_read( f"{args.shm_prefix}s_Zsmooth{args.shm_suffix}", (s_len,), dtype )[ok] s_coords = np_shm_read( f"{args.shm_prefix}s_coords{args.shm_suffix}", (s_len, 3), dtype )[ok] s_hsml = np_shm_read( f"{args.shm_prefix}s_hsml{args.shm_suffix}", (s_len,), dtype )[ok] s_oxygen = np_shm_read( f"{args.shm_prefix}s_oxygen{args.shm_suffix}", (s_len,), dtype )[ok] s_hydrogen = np_shm_read( f"{args.shm_prefix}s_hydrogen{args.shm_suffix}", (s_len,), dtype )[ok] # read in required gas particle shared memory arrays g_grpno = np_shm_read( f"{args.shm_prefix}g_grpno{args.shm_suffix}", (g_len,), "int32" ) g_sgrpno = np_shm_read( f"{args.shm_prefix}g_sgrpno{args.shm_suffix}", (g_len,), "int32" ) # Create mask for particles in the current chunk ok = np.in1d(g_sgrpno, sgrpno) * np.in1d(g_grpno, grpno) g_sgrpno = g_sgrpno[ok] g_grpno = g_grpno[ok] g_masses = np_shm_read( f"{args.shm_prefix}g_masses{args.shm_suffix}", (g_len,), dtype )[ok] g_sfr = np_shm_read( f"{args.shm_prefix}g_sfr{args.shm_suffix}", (g_len,), dtype )[ok] g_Zsmooth = np_shm_read( f"{args.shm_prefix}g_Zsmooth{args.shm_suffix}", (g_len,), dtype )[ok] g_coords = np_shm_read( f"{args.shm_prefix}g_coords{args.shm_suffix}", (g_len, 3), dtype )[ok] g_hsml = np_shm_read( f"{args.shm_prefix}g_hsml{args.shm_suffix}", (g_len,), dtype )[ok] _f = partial( assign_galaxy_prop, zed=zed, boxl=boxl, aperture=args.aperture, grpno=grpno, sgrpno=sgrpno, cop=cop, s_grpno=s_grpno, s_sgrpno=s_sgrpno, s_imasses=s_imasses, s_masses=s_masses, s_ages=s_ages, s_Zsmooth=s_Zsmooth, s_coords=s_coords, s_hsml=s_hsml, s_oxygen=s_oxygen, s_hydrogen=s_hydrogen, g_grpno=g_grpno, g_sgrpno=g_sgrpno, g_masses=g_masses, g_Zsmooth=g_Zsmooth, g_sfr=g_sfr, g_coords=g_coords, g_hsml=g_hsml, verbose=verbose, ) if numThreads == 1: pool = schwimmbad.SerialPool() elif numThreads == -1: pool = schwimmbad.MultiPool() else: pool = schwimmbad.MultiPool(processes=numThreads) galaxies = list(pool.map(_f, np.arange(len(grpno)))) pool.close() return galaxies
[docs] def np_shm_read(name: str, array_shape: tuple, dtype: str) -> NDArray[Any]: """ Read the required numpy memmap array Arguments: name (str) location of the memmap array_shape (tuple) shape of the memmap array dtype (str) data type of the memmap array Returns: Required eagle array from numpy memmap location """ tmp = np.memmap(name, dtype=dtype, mode="r", shape=array_shape) return tmp
[docs] def get_files(fileType: str, directory: str, tag: str) -> List[str]: """ Fetch filename for the different eagle outputs for reading Arguments: fileType (string) type of file in the eagle outputs directory (string) eagle data file location tag (string): snapshot tag to load Returns: array of files to read """ if fileType in ["FOF", "FOF_PARTICLES"]: files = glob.glob( "%s/groups_%s/group_tab_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SNIP_FOF", "SNIP_FOF_PARTICLES"]: files = glob.glob( "%s/groups_snip_%s/group_snip_tab_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SUBFIND", "SUBFIND_GROUP", "SUBFIND_IDS"]: files = glob.glob( "%s/groups_%s/eagle_subfind_tab_%s*.hdf5" % (directory, tag, tag) ) elif fileType in [ "SNIP_SUBFIND", "SNIP_SUBFIND_GROUP", "SNIP_SUBFIND_IDS", ]: files = glob.glob( "%s/groups_snip_%s/eagle_subfind_snip_tab_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SNIP", "SNIPSHOT"]: files = glob.glob( "%s/snipshot_%s/snip_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SNAP", "SNAPSHOT"]: files = glob.glob( "%s/snapshot_%s/snap_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["PARTDATA"]: files = glob.glob( "%s/particledata_%s/eagle_subfind_particles_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SNIP_PARTDATA"]: files = glob.glob( "%s/particledata_snip_%s/eagle_subfind_snip_particles_%s*.hdf5" % (directory, tag, tag) ) elif fileType in ["SUBFIND_PHOTOMETRY", "PHOTOMETRY"]: files = glob.glob( "%s/photometry_%s/eagle_subfind_photometry_%s*.hdf5" % (directory, tag, tag) ) else: raise ValueError("Type of files not supported") return sorted(files, key=lambda x: int(re.findall(r"(\d+)", x)[-2]))
[docs] def read_hdf5(filename: str, dataset: str) -> NDArray[Any]: """ Read the required dataset from the eagle hdf5 file Arguments: filename (str) name of the hdf5 file dataset (str) name of the dataset to extract Returns: Numpy array of the required hdf5 dataset """ with h5py.File(filename, "r") as hf: dat = np.array(hf.get(dataset)) # return empty array if empty dataset if dat.ndim == 0: return np.array([]) return dat
[docs] def read_array( ftype: str, directory: str, tag: str, dataset: str, numThreads: int = 1, noH: bool = False, physicalUnits: bool = False, CGS: bool = False, photUnits: bool = False, verbose: bool = True, ) -> Union[NDArray[Any], unyt_array[unyt_quantity]]: """ Arguments: ftype (str) eagle file type to read directory (str) location of the eagle simulation directory tag (str) snapshot tag to read dataset (str) name of the dataset to read numThreads (int) number threads to use noH (bool) remove any reduced Hubble factors physicalUnits (bool) return in physical units CGS (bool) return in CGS units verbose (bool) verbose condition Returns: Numpy/unyt array from eagle hdf5 filetype """ start = timeit.default_timer() files = get_files(ftype, directory, tag) if numThreads == 1: pool = schwimmbad.SerialPool() elif numThreads == -1: pool = schwimmbad.MultiPool() else: pool = schwimmbad.MultiPool(processes=numThreads) lg = partial(read_hdf5, dataset=dataset) dat = list(pool.map(lg, files)) # get indices of non-empty arrays non_empty_file_indices = [i for i, d in enumerate(dat) if d.shape[0] != 0] # ignore files with no data dat = [d for d in dat if d.shape[0] != 0] dat = np.concatenate(dat, axis=0) pool.close() stop = timeit.default_timer() if verbose: print( "Reading in '{}' for z = {} using {} thread(s) took {}s".format( dataset, float(tag[5:].replace("p", ".")), numThreads, np.round(stop - start, 6), ) ) if noH: dat = apply_hfreeUnits_conversion( files[non_empty_file_indices[0]], dataset, dat, verbose=verbose ) if physicalUnits: dat = apply_physicalUnits_conversion( files[non_empty_file_indices[0]], dataset, dat, verbose=verbose ) if CGS: dat = apply_CGSUnits_conversion( files[non_empty_file_indices[0]], dataset, dat, verbose=verbose ) if photUnits: dat = apply_PhotUnits( files[non_empty_file_indices[0]], dataset, dat, verbose=verbose ) return dat
[docs] def read_header( ftype: str, directory: str, tag: str, dataset: str, ) -> NDArray[Any]: """ ftype (str) eagle file type to read directory (str) location of the eagle simulation directory tag (str) snapshot tag to read dataset (str) name of the dataset to read Returns: Reads in required eagle hdf5 header data """ files = get_files(ftype, directory, tag) with h5py.File(files[0], "r") as hf: hdr = hf["Header"].attrs[dataset] return hdr
[docs] def apply_physicalUnits_conversion( filename: str, dataset: str, dat: NDArray[Any], verbose: bool = True ) -> NDArray[Any]: """ Arguments: filename (str) filename to read from dataset (str) dataset to read attribute dat (array) dataset array to apply conversion Returns: Numpy array of the dataset converted to physical units """ with h5py.File(filename, "r") as hf: exponent = hf[dataset].attrs["aexp-scale-exponent"] a = hf["Header"].attrs["ExpansionFactor"] if exponent != 0.0: if verbose: print( "Converting to physical units." "(Multiplication by a^%f, a=%f)" % (exponent, a) ) return dat * pow(a, exponent) else: if verbose: print("Converting to physical units. No conversion needed!") return dat
[docs] def apply_hfreeUnits_conversion( filename: str, dataset: str, dat: NDArray[Any], verbose: bool = True ) -> NDArray[Any]: """ Arguments: filename (str) filename to read from dataset (str) dataset to read attribute dat (array) dataset array to apply conversion Returns: Numpy array of the dataset converted to `h` free units """ with h5py.File(filename, "r") as hf: exponent = hf[dataset].attrs["h-scale-exponent"] h = hf["Header"].attrs["HubbleParam"] if exponent != 0.0: if verbose: print( "Converting to h-free units." "(Multiplication by h^%f, h=%f)" % (exponent, h) ) return dat * pow(h, exponent) else: if verbose: print("Converting to h-free units. No conversion needed!") return dat
[docs] def apply_CGSUnits_conversion( filename: str, dataset: str, dat: NDArray[Any], verbose: bool = True ) -> NDArray[Any]: """ Arguments: filename (str) filename to read from dataset (str) dataset to read attribute dat (array) dataset array to apply conversion Returns: Numpy array of the dataset converted to CGS units """ with h5py.File(filename, "r") as hf: cgs = hf[dataset].attrs["CGSConversionFactor"] if cgs != 1.0: if verbose: print( "Converting to CGS units." "(Multiplication by CGS conversion factor = %f)" % cgs ) return dat * cgs else: if verbose: print("Converting to CGS units. No conversion needed!") return dat
[docs] def apply_PhotUnits( filename: str, dataset: str, dat: NDArray[Any], verbose: bool = True ) -> unyt_array[unyt_quantity]: """ " Arguments: filename (str) filename to read from dataset (str) dataset to read attribute dat (array) dataset array to apply conversion Returns: Unyt array of the dataset in the photometric units """ with h5py.File(filename, "r") as hf: unit = hf[dataset].attrs["Units"] return unyt_array(dat, unit)
[docs] def get_star_formation_time(scale_factor: float) -> float: """ Function to convert scale factor to z Arguments: scale_factor (float) scale factor of the star particle Returns: age of the star particle in years """ SFz = (1.0 / scale_factor) - 1.0 return cosmo.age(SFz).value
[docs] def get_age( scale_factors: NDArray[Any], z: float, args: namedtuple, numThreads: int = 4, ) -> NDArray[Any]: """ Function to convert scale factor to z Arguments: scale_factors (array) scale factor of the star particle z (float) redshift args (namedtuple): parser arguments passed on to this job numThreads (int) number of threads to use Returns: array of ages of the star particles in years """ if args.age_lookup: table = np.load(args.age_lookup) scale_factor = table["scale_factor"] ages = table["ages"] Age = cosmo.age(z).value - lookup_age( scale_factors, scale_factor, ages ) else: if numThreads == 1: pool = schwimmbad.SerialPool() elif numThreads == -1: pool = schwimmbad.MultiPool() else: pool = schwimmbad.MultiPool(processes=numThreads) Age = cosmo.age(z).value - np.array( list(pool.map(get_star_formation_time, scale_factors)) ) pool.close() return Age
[docs] def assign_galaxy_prop( ii: int, zed: float, boxl: float, aperture: float, grpno: NDArray[np.int32], sgrpno: NDArray[np.int32], cop: NDArray[np.float32], s_grpno: NDArray[np.int32], s_sgrpno: NDArray[np.int32], s_imasses: NDArray[np.float32], s_masses: NDArray[np.float32], s_ages: NDArray[np.float32], s_Zsmooth: NDArray[np.float32], s_coords: NDArray[np.float32], s_hsml: NDArray[np.float32], s_oxygen: NDArray[np.float32], s_hydrogen: NDArray[np.float32], g_grpno: NDArray[np.int32], g_sgrpno: NDArray[np.int32], g_masses: NDArray[np.float32], g_Zsmooth: NDArray[np.float32], g_sfr: NDArray[np.float32], g_coords: NDArray[np.float32], g_hsml: NDArray[np.float32], verbose: bool, s_kwargs: Dict = {}, g_kwargs: Dict = {}, ) -> Galaxy: """ Load stellar and gas particle data into synthesizer galaxy object. Arguments: ii (int) galaxy number zed (float) redshift boxl (float) simulation box side length aperture (float) aperture to use from centre of potential grpno (array) Group numbers in this chunk sgrpno (array) Subgroup numbers in this chunk cop (array) centre of potential of subhalos in this chunk s_grpno (array) Stellar particle group numbers s_sgrpno (array) Stellar particle subgroup numbers s_imasses (array) Stellar particle initial masses in Msun s_masses (array) Stellar particle current masses in Msun s_ages (array) Stellar particle ages in Gyr s_Zsmooth (array) Stellar particle smoothed metallicity s_coords (array) Stellar particle coordinates in pMpc s_hsml(array) Stellar particle smoothing length in pMpc s_oxygen (array) Stellar particle abundance in oxygen s_hydrogen (array) Stellar particle abundance in hydrogen g_grpno (array) Gas particle group number g_sgrpno (array) Gas particle subgroup number g_masses (array) Gas particle masses in Msun g_Zsmooth (array) Gas particle smoothed metallicity g_sfr (array) Gas particle instantaneous SFR in Msun/yr g_coords: (array) Gas particle coordinates in pMpc g_hsml(array) Gas particle smoothing length in pMpc verbose (bool) Are we talking? s_kwargs (dictionary) kwargs for stars g_kwargs (dictionary) kwargs for gas Returns: Galaxy synthesizer galaxy object """ galaxy = Galaxy(redshift=zed, verbose=verbose) # set the bounds of the box length bounds = np.array([boxl, boxl, boxl]) # Fill individual galaxy objects with star particles # mask for current galaxy ok = np.where((s_grpno == grpno[ii]) * (s_sgrpno == sgrpno[ii]))[0] # correcting for periodic boundary r = cop[ii] - s_coords[ok] r = np.sign(r) * np.min(np.dstack(((r) % bounds, (-r) % bounds)), axis=2) # mask for aperture ok = ok[norm(r, axis=1) <= aperture] # Assign stellar properties galaxy.load_stars( initial_masses=s_imasses[ok] * Msun, current_masses=s_masses[ok] * Msun, ages=s_ages[ok] * 1e9 * yr, metallicities=s_Zsmooth[ok], coordinates=s_coords[ok] * Mpc, smoothing_lengths=s_hsml[ok] * Mpc, s_oxygen=s_oxygen[ok], s_hydrogen=s_hydrogen[ok], **s_kwargs, ) # Fill individual galaxy objects with gas particles sfr_flag = g_sfr > 0 # mask for current galaxy ok = np.where((g_grpno == grpno[ii]) * (g_sgrpno == sgrpno[ii]))[0] # correcting for periodic boundary r = cop[ii] - g_coords[ok] r = np.sign(r) * np.min(np.dstack(((r) % bounds, (-r) % bounds)), axis=2) # # mask for aperture ok = ok[norm(r, axis=1) <= aperture] # Assign gas particle properties galaxy.load_gas( masses=g_masses[ok] * Msun, metallicities=g_Zsmooth[ok], star_forming=sfr_flag[ok], coordinates=g_coords[ok] * Mpc, smoothing_lengths=g_hsml[ok] * Mpc, **g_kwargs, ) galaxy.gas.dust_to_metal_ratio = galaxy.dust_to_metal_vijayan19() return galaxy