slicersim

Simulation of Slicer observations

---

reference

See the dedicated paper: Rigault et al. (2026)

---

Installation

pip install slicersim

---

Top level ETC

>>> import slicersim
>>> import numpy as np

Any spectra

Compute the exposure time needed any input spectrum:

>>> lbda_ref = np.linspace(3000, 20_000, 500) # larger than lazuli bandpass
>>> flux_ref = np.ones(lbda_ref.shape) # flat spectrum
>>> exptime, lazulitarget = slicersim.lazuli_etc(lbda_ref, flux_ref, snr=20, mag=21, band="bessellb")
>>> print(exptime)
237.72

Supernova

Compute the exposure time needed to observe a Supernovae:

>>> exptime, snia_target = slicersim.lazuli_sn_etc(snr=20, model="salt", redshift=1.2, x1=1.5, c=0.2, phase=-2.2)
>>> print(exptime)
5795.84

---

Quick look

import slicersim
# load a SN Ia (see also: `slicersim.LazuliTarget(lbda, flux)`)
snia = slicersim.LazuliSN(redshift=0.8, c=0.2, x1=-1.2, phase=1.5)

# update configuration to reach a signal-to-noise of 20
_ = snia.setup_to_snr(20)

# grab the expected observed spectrum
lbda, flux_1, variance_1 = snia.get_spectrum(unit="adu")

# change the target property
# warning: without updating the setting, this new target won't have a snr=20
snia.set_properties(redshift=1.2)
lbda, flux_2, variance_2 = snia.get_spectrum(unit="adu")

and show your simulated spectra

import matplotlib.pyplot as plt
import numpy as np
fig, ax = plt.subplots(figsize=[7,3])

ax.plot(lbda, flux_1)
ax.fill_between(lbda, 
                flux_1-np.sqrt(variance_1),
                flux_1+np.sqrt(variance_1), alpha=0.3,
               label="z=0.8")

ax.plot(lbda, flux_2)
ax.fill_between(lbda, 
                flux_2-np.sqrt(variance_2),
                flux_2+np.sqrt(variance_2), alpha=0.3,
               label="z=1.2")
ax.legend(frameon=False, fontsize="small")
ax.set(xlabel=r"wavelength [$\AA$]", ylabel="flux [ADU]")

---

Tutorials

Beginner

• ETC and simuated spectra: top level function
  • Exposure time calculator for any target or a supernovae
  • get the simulated spectrum and variance
• LazuliTarget (Any target, Type Ia Supernovae, CalSpec):
  • Specify the desired signal to noise
  • get exposure time, and read-mode
  • get simulated spectra
• Check the origin of the variance sources
  • origin of variance (read-out noise, target poisson noise, dark-current etc)
  • switch off any contribution and see resulting variance
  • scan all variance contributions, get the resulting dataframe and plot the result.
• Change properties of the target:
  • change any supernovae properties
  • change the magnitude of a loaded target

Advanced

• Change the detector read-out mode or spectrograph spatial sampling
  • change the detector mode: max-group, n-frames per group
  • force the read-our mode (and see which SNR you eventually get)
  • change the spectrograph sampling (fine and medium grid)
• Access the detector QE, throughtput, effective spectral resolution...
• Access any property of any simulation element, and change them !

Experts

• Noise Equivalent area
• Build you own “scene”
• Build your own configuration

---

Lower-level: Simulator & config.

Example: Study origin of variance

You can load a slicersim object at a lower level than an lazuli object: a Simulator. This object is at the core of slicersim and combines information of all sources needed to simulate our observations: a scene (what is observed), a mirror, a spectrograph and a detector.

The Simulator combine these to be able to generate a datacube with realistic noise, and extract a realistic spectrum from it.

Once your simulator is loaded, you have several method to check the variance origin.

• estimate_variance_contribution: that probe the variance origin for a given wavelength rate
• estimate_variance_contribution_spectra: similar but for whole spectrum
• show_variance_sources: plotting function associated to estimate_variance_contribution_spectra

import slicersim
# load the correct simulator
config = slicersim.iotools.get_config(instrument='lazuli.toml')
sim = slicersim.Simulation.from_config(config)

# Set the target you want
sim.update(target__redshift=1., target__x1=0, target__c=0)

# look for the config needed to get an average SNR of 20 in [4000, 6800] rest-frame
new_config, snr, integration_time = sim.fetch_snr(20, lbda_range=[4000, 6800], frame='rest')
sim.update(**new_config) # let's set it

# get the expected spectrum (in ADU)
lbda, flux, variance = sim.get_spectrum()

# Show details
fig = sim.show_variance_sources(flux_calibrated=False)

---

PSF & Noise equivalent area

by default, the code assumes gaussian PSF both for spatial (at the slicer/mla) and cross-dispersion (at pixels) level.

slicersim.profiles contains additional PSF model (from astropy.modeling that can be used to build 2D model or estimate the PSF noise equivalent area (nea).

(more to come..., see: slicersim.profiles.get_2dpsf_nea())

---

Credits

Developped by M. Rigault ; adapted from the original MLAPerf v:0.18.0 developed by Y. Copin and M. Rigault