autoprof

name	autoprof
description	AutoProf non-parametric galaxy isophote fitting pipeline. Use when fitting galaxy surface brightness profiles with AutoProf's FFT-based method, benchmarking against photutils/GALFIT, or running AutoProf programmatically on FITS images.

AutoProf

Non-parametric isophotal analysis pipeline for galaxy images (Stone et al. 2021). Uses FFT-based fit-stabilization to robustly extract SB profiles ~2 mag/arcsec² fainter than photutils on the same images.

• Docs: https://autoprof.readthedocs.io/en/latest/
• Code: https://github.com/Autostronomy/AutoProf
• Paper: https://arxiv.org/abs/2106.13809
• Install: pip install autoprof (v1.3.4)
• Successor: AstroPhot (parametric, multi-band) — use for PSF-limited or crowded fields

Critical: Environment Isolation Required

AutoProf v1.2+ pins photutils<=1.5.0, which conflicts with modern environments. Always use a separate virtual environment:

python3 -m venv /path/to/autoprof_venv
/path/to/autoprof_venv/bin/pip install autoprof==1.3.4

Call via /path/to/autoprof_venv/bin/python script.py.

Additional constraints:

• Input FITS must be uncompressed (.fits, not .fits.fz). Decompress first.
• AutoProf reads image from HDU 0 by default; use ap_image_hdu to change.

Programmatic Usage

Pipeline approach (recommended)

from autoprof.Pipeline import Isophote_Pipeline

pipeline = Isophote_Pipeline(loggername="autoprof.log")
result = pipeline.Process_Image(options={
    "ap_image_file": "/path/to/galaxy.fits",  # uncompressed FITS
    "ap_mask_file": "/path/to/mask.fits",     # optional; 0=good, >0=bad
    "ap_mask_hdu": 0,
    "ap_name": "galaxy_name",
    "ap_pixscale": 0.262,       # arcsec/pixel (required)
    "ap_zeropoint": 22.5,       # nanomaggy zeropoint
    "ap_process_mode": "image",
    "ap_doplot": False,
    "ap_saveto": "/output/dir/",
    "ap_isoclip": True,         # sigma-clip along isophotes
})
# result is dict of step timings if success, or 1 if failure

Direct function calls (bypass pipeline, access results in-memory)

from autoprof.pipeline_steps import (
    Background_Mode, PSF_Assumed, Center_HillClimb,
    Isophote_Initialize, Isophote_Fit_FFT_Robust,
    Isophote_Extract, EllipseModel,
)

img = fits.getdata("galaxy.fits").astype(float)
options = {
    "ap_pixscale": 0.262, "ap_zeropoint": 22.5,
    "ap_name": "test", "ap_isoclip": True, "ap_doplot": False,
}
results = {}

for step in [Background_Mode, PSF_Assumed, Center_HillClimb,
             Isophote_Initialize, Isophote_Fit_FFT_Robust,
             Isophote_Extract]:
    img, res = step(img, results, options)
    results.update(res)

# Profile in results["prof data"]
# Model image via EllipseModel step

Output Format

Profile columns (.prof file, CSV)

The .prof file has two header lines: line 0 is # units... (comment), line 1 is column names. Parse with: np.genfromtxt(file, delimiter=",", names=True, skip_header=1)

Column	Unit	Description
R	arcsec	Semi-major axis (in arcsec, not pixels)
SB	mag/arcsec²	Surface brightness (99.999 = invalid/below noise)
SB_e	mag/arcsec²	SB error
totmag	mag	Cumulative enclosed magnitude
totmag_e	mag	Enclosed magnitude error
ellip	—	Ellipticity (1 - b/a)
ellip_e	—	Ellipticity error
pa	deg	Position angle (astronomical: CCW from +y/North)
pa_e	deg	PA error
pixels	count	Pixels sampled
maskedpixels	count	Masked pixels
totmag_direct	mag	Direct-sum enclosed magnitude

Auxiliary file (.aux, YAML-like)

Contains: center (x, y), background level, PSF FWHM, initial geometry, fit flags, processing time.

Key Parameters

Parameter	Default	Purpose
ap_pixscale	(required)	Pixel scale (arcsec/pixel)
ap_zeropoint	22.5	Photometric zero point
ap_isoclip	False	Sigma-clip along isophotes
ap_isoclip_nsigma	5	Clipping threshold
ap_fit_limit	2	Stop at N × background noise
ap_samplegeometricscale	0.1	Radial spacing growth factor
ap_isofit_iterlimitmax	300	Max fitting iterations
ap_isoaverage_method	"median"	Averaging along isophotes
ap_fluxunits	"mag"	Output: "mag" or "intensity"
ap_iso_measurecoefs	None	Fourier modes to measure, e.g. (3,4)
ap_isoinit_pa_set	None	Override initial PA (deg, astro convention)
ap_isoinit_ellip_set	None	Override initial ellipticity
ap_set_center	None	Override center: {"x": float, "y": float}

Pipeline Steps

Default order: background → psf → center → isophoteinit → isophotefit → isophoteextract → checkfit → writeprof

The core fitter is Isophote_Fit_FFT_Robust — FFT-based with regularization, distinct from photutils' Jedrzejewski (1987) iterative method.

Conventions

• PA: astronomical convention (CCW from +y/North), same as SGA
• SMA: output R is in arcsec (R_pixels × ap_pixscale)
• SB: already in mag/arcsec² (no conversion needed for comparison)
• Mask: 0 = good, nonzero = bad (same convention as our pipeline masks)
• SB=99.999: sentinel for invalid/below-noise isophotes — filter these out

Comparison with photutils

Feature	AutoProf	photutils
Algorithm	FFT-based + regularization	Jedrzejewski (1987) iterative
PA convention	Astronomical (CCW from +y)	Math (CCW from +x)
SMA output unit	arcsec	pixels
SB output	mag/arcsec²	raw intensity (nanomaggies/pixel)
Harmonics	via ap_iso_measurecoefs	always measured (a3/b3/a4/b4)
Center	Auto hill-climb or fixed	Via EllipseGeometry
Background	Auto-estimated	Not handled (user subtracts)
Model image	EllipseModel step	build_ellipse_model()
Env constraint	Needs isolated venv	Works with modern stack