Skip to main content

PyAutoCTI: Charge Transfer Inefficiency Modeling

Documentation Status Tests Build code-style arXiv

Installation Guide | readthedocs | What is CTI?

Charge Transfer Inefficiency, or CTI for short, is an effect that occurs when acquiring imaging data from Charge Coupled Devices (CCDs). Due to radiation damage to the CCD’s silicon lattice electrons are read-out inefficiently, creating a characteristic trailing or smearing effect.

Here is an example of CTI in the Hubble space telescope, after decades of radiation damage:

Alternative text

PyAutoCTI makes it simple to calibrate a time-varying CTI model using in-orbit observations and correct CTI in science imaging using this model.

PyAutoCTI development is centred around mitigating CTI for the Euclid space mission, which relies on the precise measurement of galaxy shapes in order to map out the distribution of dark matter throughout the Universe via a phenomena called weak gravitational lensing.

Getting Started

The following links are useful for new starters:

API Overview

To model CTI, PyAutoCTI wraps the library arCTIc (https://github.com/jkeger/arctic).

CTI can be added to an image as follows:

import autocti as ac

"""
Define a pre-cti image which **PyAutoCTI** adds CTI to.
"""
pre_cti_data_2d = ac.Array2D.no_mask(
            values=[
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 10.0, 10.0, 10.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 10.0, 10.0, 10.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 10.0, 10.0, 10.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
        [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
    ],
    pixel_scales=0.1,
)

"""
A clocker object is used to model CCD clocking, which includes customization such
as the properties of the read-out electronics.
"""
clocker_2d = ac.Clocker2D(parallel_roe=ac.ROE())

"""
CTI is caused by traps on the CCD's silicon lattice, for example traps which capture
electrons instantaneously.
"""
parallel_trap = ac.TrapInstantCapture(density=100.0, release_timescale=1.0)

"""
CTI also depends on how electrons fill each pixel in a CCD, therefore we define
the volume-filling properties of the CCD.
"""
parallel_ccd = ac.CCDPhase(
    well_fill_power=0.58, well_notch_depth=0.0, full_well_depth=200000.0
)

"""
The data, traps and CCD properties are combined to clock the pre-CTI data and return the
post-CTI data.
"""
post_cti_data_2d = clocker_2d.add_cti(
    data=pre_cti_data_2d,
    parallel_trap_list=[parallel_trap],
    parallel_ccd=parallel_ccd
)

"""
We can use PyAutoCTI's built in visualization library to plot the data with CTI.
"""
import autocti.plot as aplt

array_2d_plotter = aplt.Array2DPlotterarray=post_cti_data_2d)
array_2d_plotter.figure_2d()

With PyAutoCTI, you can begin calibrating a CTI model in minutes. The example below demonstrates a simple analysis which fits a CTI model to charge injection imaging calibrate data (a form of data used to calibrate a CTI model)

import autofit as af
import autocti as al
import autocti.plot as aplt

"""
Define the 2D shape of the charge injection image.
"""
shape_native = (30, 30)

"""
Define where the charge injection is on the data.
"""
regions_list = [(0, 25, serial_prescan[3], serial_overscan[2])]

"""
Setup the data layout which informs **PyAutoCTI** where information on
CTI is in the data.
"""
layout = ac.Layout2DCI(
    shape_2d=shape_native,
    region_list=regions_list,
)

"""
Load the charge injection image from fits.
"""
dataset = ac.ImagingCI.from_fits(
    data_path=path.join(dataset_path, f"data.fits"),
    noise_map_path=path.join(dataset_path, f"noise_map.fits"),
    pre_cti_data_path=path.join(dataset_path, f"pre_cti_data.fits"),
    layout=layout,
    pixel_scales=0.1,
)

"""
Again define the clocker which models CCD clocking and read-out electronics.
"""
clocker_2d = ac.Clocker2D(parallel_roe=ac.ROE())

"""
Define the traps in the CTI model and customize the priors of their free parameters.
"""
trap = af.Model(ac.TrapInstantCapture)

trap.density = af.UniformPrior(lower_limit=0.0, upper_limit=20.0)
trap.release_timescale = af.UniformPrior(lower_limit=0.0, upper_limit=20.0)

"""
Define the CCD filling behaviour of the CTI, which is also part of the model and is
fitted for as free parameters.
"""
parallel_ccd = af.Model(ac.CCDPhase)

parallel_ccd.well_fill_power = af.UniformPrior(lower_limit=0.0, upper_limit=1.0)
parallel_ccd.well_notch_depth = 0.0
parallel_ccd.full_well_depth = 200000.0

"""
We define the non-linear search used to fit the model to the data (in this case, Dynesty).
"""
search = af.Nautilus(name="search[example]", n_live=50)

"""
We next set up the `Analysis`, which contains the `log likelihood function` that the
non-linear search calls to fit the cti model to the data.
"""
analysis = ac.AnalysisImagingCI(dataset=dataset, clocker=clocker_2d)

"""
To perform the model-fit we pass the model and analysis to the search's fit method. This will
output results (e.g., dynesty samples, model parameters, visualization) to hard-disk.
"""
result = search.fit(model=model, analysis=analysis)

"""
The results contain information on the fit, for example the maximum likelihood
model from the Dynesty parameter space search.
"""
print(result.samples.max_log_likelihood())

Support

Support for installation issues, help with cti modeling and using PyAutoCTI is available by raising an issue on the GitHub issues page.

We also offer support on the PyAutoCTI Slack channel, where we also provide the latest updates on PyAutoCTI. Slack is invitation-only, so if you’d like to join send an email requesting an invite.

Metadata

Release files for autocti 2026.11.6.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for autocti 2026.11.6.1
File Size Uploaded Provenance Actions
autocti-2026.11.6.1.tar.gz 4.5 MB Without provenance Details for autocti-2026.11.6.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for autocti 2026.11.6.1
File Interpreter ABI Platform Actions
autocti-2026.11.6.1-py3-none-any.whl Python 3 none any Details for autocti-2026.11.6.1-py3-none-any.whl

Total release size: 4.7 MB

Release files / autocti-2026.11.6.1.tar.gz

Download URL autocti-2026.11.6.1.tar.gz
Size 4.5 MB
Tags Source
SHA-256 checksum
How to use checksums
62ca58808ccf291bfe45e3194730861c1b08616d4610500acabe30840bf088e1
BLAKE2b-256 checksum
How to use checksums
49c35bd07e4e73fe51e72098e939e34a7517b5b137a715817d3067c576b64e86
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.0.1 CPython/3.12.16

Release files / autocti-2026.11.6.1-py3-none-any.whl

Download URL autocti-2026.11.6.1-py3-none-any.whl
Size 171.0 kB
Tags Python 3
SHA-256 checksum
How to use checksums
5cacd56d7980956005cc3226035022d6017e24d8da5ab46499a217e36cc02490
BLAKE2b-256 checksum
How to use checksums
dd45de5c620daead0b60a9c1c13bfc2590f79dd49a5483618b8a9dfab879c256
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.0.1 CPython/3.12.16

Security for autocti 2026.11.6.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release
Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page