specwizard 0.9.8.0

.

Overview

SPECWIZARD is a Python package to compute and analyze mock quasar absorption spectra from cosmological simulations.

This repository provides:

• Tools to build and validate simulation input dictionaries (“Wizard” dictionaries).

• Routines to generate short spectra for single sightlines.

• Routines to assemble full long spectra across a redshift path.

• HDF5 I/O helpers to save and read spectra in a consistent format.

Citing SPECWIZARD

If you use SPECWIZARD in scientific work, please cite the relevant code/paper for your release.

Installation Notes

SPECWIZARD is written in python3 and stable versions are available on PyPI. The easiest installation method is:

pip install specwizard

The code is constantly updated. Feedback and bug reports are welcome through email or GitHub.

Quick Start

Most users follow this flow:

1. Create a YAML configuration file.

2. Load it with Build_Input.read_from_yml.

3. Generate spectra.

4. Save to HDF5.

5. Read the file back for analysis.

Minimal Configuration (Wizard YAML)

Save a file such as Wizard.yml:

file_type:
    sim_type: swift
    snap_type: snapshot

snapshot_params:
    directory: /path/to/snapshots
    file: snapshot_012.hdf5

sightline:
    ProjectionAxes: [simx, simy, simz]
    ProjectionStart: [0.5, 0.5, 0.0]
    ProjectionLength: 1
    SightLength: null
    ProjectionExtend:
        extend: false
        extendfactor: 3
    nsight: 0

ionparams:
    table_type: specwizard_cloudy
    iondir: /path/to/ion_tables
    fname: null
    ions:
        - [Hydrogen, H I]
        - [Carbon, C IV]
    SFR_properties:
        modify_particle: true
        ignore_particle: false
        Temperature [K]: 1.0E+4
    atomfile: /path/to/atomic_info.hdf5

ODParams:
    VelOffset_kms: 0
    PecVelEffectsOff: false
    ThermalEffectsOff: false
    VoigtOff: false

extraparams:
    periodic: true
    pixkms: 1
    ReadIonFrac:
        ReadIonFrac: false
        ReadHydrogen: true
        HI: NeutralHydrogenAbundance
        ReadHelium: false
        He: ""
        fname_urchin: ""

Output:
    directory: ./outputs/
    fname: shortspec.hdf5

Notes:

• Build_Input.read_from_yml reads file_type, snapshot_params, sightline, ionparams, ODParams and optional LongSpectra, extraparams, and Output sections.

• Output is required if you plan to use OpticalDepth_IO to save data.

Generate And Save A Short Spectrum

import specwizard as spw

# 1) Build Wizard dictionary from YAML
builder = spw.Build_Input()
wizard = builder.read_from_yml("Wizard.yml")

# 2) Generate short-spectrum products
opticaldepth, projected_los, particles = spw.GenerateShortSpectra(wizard)

# 3) Save to HDF5 using the package writer
io = spw.OpticalDepth_IO(wizard=wizard, create=True)
payload = {
        "nsight": wizard["sightline"]["nsight"],
        "Projection": projected_los,
        "OpticaldepthWeighted": opticaldepth,
}
io.write_shortspectra_to_file(payload)

Read A Saved Short Spectrum

import specwizard as spw

builder = spw.Build_Input()
wizard = builder.read_from_yml("Wizard.yml")

io = spw.OpticalDepth_IO(wizard=wizard, create=False)
short_data = io.read_shortspectra_from_file()

# Examples of access:
header = short_data["Header"]
one_los = next(iter(short_data["Data"]))
los_data = short_data["Data"][one_los]

Generate, Save, And Read A Long Spectrum

To generate a long spectrum, include a LongSpectra block in your YAML:

LongSpectra:
    lambda_min: 945.0
    lambda_max: 8000.0
    dlambda: 0.5
    z_qsr: 3.0
    delta_z: 0.01
    all_contaminants: false
    file_dir: /path/to/los/files/

Then run:

import specwizard as spw

builder = spw.Build_Input()
wizard = builder.read_from_yml("Wizard.yml")

long_builder = spw.LongSpectra(wizard)
coven, redshifts = long_builder.create_coven()
long_spectra = long_builder.do_long_spectra(coven)

# Optional post-processing
# long_spectra = long_builder.add_contaminants(long_spectra)
# long_spectra = long_builder.add_HI_damping_wings(long_spectra, n=2)

# Save
io = spw.OpticalDepth_IO(wizard=wizard, create=True)
io.write_fullspectrum_to_file(long_spectra)

# Read
io_read = spw.OpticalDepth_IO(wizard=wizard, create=False)
long_data = io_read.read_fullspectrum_from_file()

Data Fields: What You Can Read

This project writes structured HDF5 groups for both short and long spectra. Below are the common fields and where to find them.

• Short spectra (per LOS):

  • Top-level groups: LOS_<n>/ and Header/.

  • Element-weighted properties: LOS_<n>/<Element>/Element-weighted/<field> (e.g. Velocities, Densities, Temperatures).

  • Ion-weighted properties: LOS_<n>/<Element>/<Ion>/Ion-weighted/<field>.

  • Optical-depth-weighted properties for each ion: LOS_<n>/<Element>/<Ion>/Ion optical depth-weighted/<field> (includes Optical depths, Velocities, Densities, Temperatures, Metallicities, HydrogenDensities where present).

  • If the simulation includes non-equilibrium / tracked ionic abundances, additional groups are written under SimIon-weighted and SimIon optical depth-weighted with the same fields as the tabulated ions.

• Long spectra (full-spectrum accumulation):

  • Global grids: FullSpectrum/Velocities and FullSpectrum/Wavelengths.

  • Per-ion data: FullSpectrum/<Element>/<Ion>/... with datasets for lambda0, f-value and fields such as Optical depths, Velocities, Densities, Temperatures, Metallicities, HydrogenDensities (each saved together with metadata/attributes when available).

  • The long-spectrum writer prefers simulation-tracked ion optical-depth-weighted fields when present: if SimIons optical-depth-weighted quantities exist they will be used (and saved) in preference to tabulated ion optical-depth-weighted values.

• Access helpers:

  • Use OpticalDepth_IO.ReadVariable(path) to read a specific dataset and its attributes.

  • OpticalDepth_IO.ReadHeader() returns header attributes and any stored global datasets.

Output And File Layout

• Output path is controlled by wizard['Output']['directory'] and wizard['Output']['fname'].

• Short-spectrum files contain LOS-based groups such as LOS_0/... plus Header/....

• Full long-spectrum files are stored under FullSpectrum/....

• OpticalDepth_IO.ReadVariable(path) can be used to load a specific dataset with metadata.