tdanalysis 1.0.0__tar.gz

tdanalysis-1.0.0/LICENSE

@@ -0,0 +1,203 @@
+tdanalysis is licensed under the PolyForm Noncommercial License 1.0.0, WITH
+TWO ADDITIONAL TERMS stated at the end of this file. It is therefore a
+MODIFIED PolyForm licence and must not be described as plain PolyForm
+Noncommercial 1.0.0, nor identified by that SPDX identifier.
+
+Copyright (c) 2026 Vaishak Prasad and Viviana Cáceres
+
+WHAT THIS MEANS IN SHORT, with the licence text below governing:
+
+  * Research, teaching, personal and other noncommercial use: permitted,
+    freely, with no need to contact anyone or pay anything.
+  * Commercial use: NOT permitted under this licence. Contact the licensor to
+    arrange separate terms. Those terms may require a fee or a royalty.
+  * Using this work as training data for machine-learning models: NOT
+    permitted under this licence, commercial or not. See ADDITIONAL TERM 1.
+  * Working on or with this software using an AI tool that retains or trains
+    on what you give it: NOT permitted. See ADDITIONAL TERM 2.
+
+This applies from the first release. No earlier version of tdanalysis has been
+published under any other licence.
+
+--------------------------------------------------------------------------
+
+# PolyForm Noncommercial License 1.0.0
+
+<https://polyformproject.org/licenses/noncommercial/1.0.0>
+
+## Acceptance
+
+In order to get any license under these terms, you must agree
+to them as both strict obligations and conditions to all
+your licenses.
+
+## Copyright License
+
+The licensor grants you a copyright license for the
+software to do everything you might do with the software
+that would otherwise infringe the licensor's copyright
+in it for any permitted purpose.  However, you may
+only distribute the software according to [Distribution
+License](#distribution-license) and make changes or new works
+based on the software according to [Changes and New Works
+License](#changes-and-new-works-license).
+
+## Distribution License
+
+The licensor grants you an additional copyright license
+to distribute copies of the software.  Your license
+to distribute covers distributing the software with
+changes and new works permitted by [Changes and New Works
+License](#changes-and-new-works-license).
+
+## Notices
+
+You must ensure that anyone who gets a copy of any part of
+the software from you also gets a copy of these terms or the
+URL for them above, as well as copies of any plain-text lines
+beginning with `Required Notice:` that the licensor provided
+with the software.  For example:
+
+> Required Notice: Copyright Yoyodyne, Inc. (http://example.com)
+
+## Changes and New Works License
+
+The licensor grants you an additional copyright license to
+make changes and new works based on the software for any
+permitted purpose.
+
+## Patent License
+
+The licensor grants you a patent license for the software that
+covers patent claims the licensor can license, or becomes able
+to license, that you would infringe by using the software.
+
+## Noncommercial Purposes
+
+Any noncommercial purpose is a permitted purpose.
+
+## Personal Uses
+
+Personal use for research, experiment, and testing for
+the benefit of public knowledge, personal study, private
+entertainment, hobby projects, amateur pursuits, or religious
+observance, without any anticipated commercial application,
+is use for a permitted purpose.
+
+## Noncommercial Organizations
+
+Use by any charitable organization, educational institution,
+public research organization, public safety or health
+organization, environmental protection organization,
+or government institution is use for a permitted purpose
+regardless of the source of funding or obligations resulting
+from the funding.
+
+## Fair Use
+
+You may have "fair use" rights for the software under the
+law. These terms do not limit them.
+
+## No Other Rights
+
+These terms do not allow you to sublicense or transfer any of
+your licenses to anyone else, or prevent the licensor from
+granting licenses to anyone else.  These terms do not imply
+any other licenses.
+
+## Patent Defense
+
+If you make any written claim that the software infringes or
+contributes to infringement of any patent, your patent license
+for the software granted under these terms ends immediately. If
+your company makes such a claim, your patent license ends
+immediately for work on behalf of your company.
+
+## Violations
+
+The first time you are notified in writing that you have
+violated any of these terms, or done anything with the software
+not covered by your licenses, your licenses can nonetheless
+continue if you come into full compliance with these terms,
+and take practical steps to correct past violations, within
+32 days of receiving notice.  Otherwise, all your licenses
+end immediately.
+
+## No Liability
+
+***As far as the law allows, the software comes as is, without
+any warranty or condition, and the licensor will not be liable
+to you for any damages arising out of these terms or the use
+or nature of the software, under any kind of legal claim.***
+
+## Definitions
+
+The **licensor** is the individual or entity offering these
+terms, and the **software** is the software the licensor makes
+available under these terms.
+
+**You** refers to the individual or entity agreeing to these
+terms.
+
+**Your company** is any legal entity, sole proprietorship,
+or other kind of organization that you work for, plus all
+organizations that have control over, are under the control of,
+or are under common control with that organization.  **Control**
+means ownership of substantially all the assets of an entity,
+or the power to direct its management and policies by vote,
+contract, or otherwise.  Control can be direct or indirect.
+
+**Your licenses** are all the licenses granted to you for the
+software under these terms.
+
+**Use** means anything you do with the software requiring one
+of your licenses.
+
+---
+
+Required Notice: Copyright (c) 2026 Vaishak Prasad and Viviana Cáceres. Commercial use requires a
+separate written licence from the copyright holder — contact vbp5193@psu.edu.
+See NOTICE.md.
+
+--------------------------------------------------------------------------
+
+## ADDITIONAL TERM 1: machine-learning training
+
+This term is added by the licensor and is not part of the published PolyForm
+Noncommercial License 1.0.0.
+
+Use of the software, its documentation, or its outputs as training,
+fine-tuning, validation or evaluation data for a machine-learning model,
+statistical language model, or other automated system that generates or
+reproduces code or text, is NOT a permitted purpose under these terms,
+whether or not that use is commercial. Contact the licensor for permission.
+
+This does not restrict using the software for its intended scientific
+purpose, including as a component of, or a source of training data for,
+scientific models of gravitational-wave signals -- for example emulators,
+surrogates or reduced-order models of the waveforms or posteriors this
+software produces. That is ordinary research use and is permitted.
+
+## ADDITIONAL TERM 2: AI tools used on this software
+
+This term is added by the licensor and is not part of the published PolyForm
+Noncommercial License 1.0.0.
+
+You may use an AI assistant, coding agent, or similar automated system to
+read, write, analyse or modify this software ONLY where that system is
+configured so that the software, its documentation, and anything you derive
+from them are not retained beyond serving your immediate request and are not
+used to train, fine-tune, or evaluate any model.
+
+Where such a tool offers that setting -- it goes by names like "training
+off", "do not use my data for model improvement", or "zero data retention"
+-- it must be enabled for the whole of your use of this software. Where a
+tool offers no such setting, or where its terms of service reserve a right
+to train on what you submit, use of that tool with this software is not
+permitted under these terms.
+
+This term does NOT prohibit AI assistance as such. It prohibits one
+configuration of it: the one that would feed this software into a model's
+training corpus, which is what ADDITIONAL TERM 1 already forbids doing
+directly. The purpose is to close the same door twice, not to bar a tool
+that keeps your input out of its training data.

tdanalysis-1.0.0/NOTICE.md

@@ -0,0 +1,133 @@
+# Notices
+
+Copyright (c) 2026 Vaishak Prasad and Viviana Cáceres. All rights reserved except as licensed below.
+
+This file summarises the licensing position. `LICENSE` governs; where the two
+differ, `LICENSE` wins.
+
+## Licence
+
+`tdanalysis` is licensed under the **PolyForm Noncommercial License 1.0.0, with
+two additional terms** on machine-learning training and on AI tooling. The full
+text is in `LICENSE`.
+
+Because those terms modify it, this is **not** plain PolyForm Noncommercial and
+must not be identified by that SPDX id. The metadata declares
+`LicenseRef-PolyForm-Noncommercial-1.0.0-with-AI-terms`.
+
+**This is not an open-source licence.** It is deliberately not OSI-approved,
+because it restricts the field of use, which the Open Source Definition forbids.
+Two consequences worth planning around: conda-forge will not accept it, and
+neither will JOSS or most software-paper venues. PyPI has no such requirement.
+
+This applies from the first release. No version of `tdanalysis` has been
+published under any other licence.
+
+## Copyright, authorship and contributors
+
+Copyright is held jointly by **Vaishak Prasad** and **Viviana Cáceres**, the two
+principal developers of `tdanalysis`.
+
+Copyright holder and contributor are different things, and this section lists
+both. Contributors to this release, and to the wider project it is drawn from:
+
+* **Vaishak Prasad** — the parameter estimation core, the consistency test, the
+  covariance and quadratic-form machinery, and the runner that this tier
+  consists of.
+* **Viviana Cáceres** — the ringdown spectroscopy and multi-segment analysis,
+  and contributions across the shared code this release builds on.
+* **Sayan Neogi** — a fix to the merger placement, where a missing `abs()` had
+  the peak located on the signed strain rather than on its amplitude.
+
+The source repository was licensed GPL-3.0 from its first commit until this
+release, and the change of licence was made with the agreement of both copyright
+holders. It is prospective: anyone who received a copy of the software under
+GPL-3.0 keeps GPL-3.0 rights to that copy. A licence already granted cannot be retracted, and
+nothing here attempts to do so.
+
+## What you may do without asking
+
+Research, teaching, academic and other noncommercial use, including publishing
+results obtained with this software. No permission, no contact, no payment. The
+licence grants this outright to "any charitable organization, educational
+institution, public research organization, public safety or health organization,
+environmental protection organization, or government institution ... regardless
+of the source of funding" — so academic use funded by industry is still
+noncommercial use, and needs nothing from anyone.
+
+Modifying it, and sharing your modifications, for those same noncommercial
+purposes.
+
+## What requires permission
+
+**Commercial use of any kind.** If you or your company would use this software
+in connection with a product, service, or activity from which commercial
+advantage or monetary compensation is anticipated, you need a separate licence.
+Commercial licences are available and separate terms can be arranged; those
+terms may require a fee or a royalty. The licence is not a refusal, it is a
+request that you ask.
+
+**Use as training data.** Using the software, its documentation, or its outputs
+to train, fine-tune, validate or evaluate a machine-learning model, language
+model, or other system that generates or reproduces code or text — commercial or
+not. See ADDITIONAL TERM 1.
+
+This restriction is **not** aimed at scientific modelling. Building an emulator,
+surrogate or reduced-order model *of the waveforms or posteriors this software
+produces* is ordinary research use and is permitted. The term is aimed at
+ingestion of the source and documentation into general-purpose code and language
+models.
+
+**AI tools configured to train on what you give them.** Reading, writing or
+modifying this software with an AI assistant or coding agent is permitted only
+where that tool is configured not to retain your input or use it for training.
+See ADDITIONAL TERM 2. This does not prohibit AI assistance; it prohibits the
+one configuration that would route the software into a training corpus by
+another path.
+
+Contact **vbp5193@psu.edu** for any of the above.
+
+## Reservation of rights for text and data mining
+
+The copyright holders **expressly reserve** the right to reproduce, extract, or
+otherwise use this software, in whole or in part, for text and data mining, or
+for the training, fine-tuning, evaluation, or validation of machine-learning or
+generative-AI models. No such right is granted by the licence above, whether the
+use would otherwise be commercial or noncommercial.
+
+This reservation is made in machine-readable form for the purposes of Article
+4(3) of Directive (EU) 2019/790, and is asserted generally elsewhere.
+
+```
+tdm-reservation: all
+ai-training: disallowed
+```
+
+**An honest note on how strong these are.** They are not equally strong, and it
+is better to know which is which.
+
+The noncommercial term rests on copyright and is the strongest of the three. A
+commercial user who ignores it is using a copyrighted work outside its licence.
+
+The training restrictions — ADDITIONAL TERM 1 and the reservation above — are
+weaker. Whether model training requires a copyright licence at all is unsettled,
+and in some jurisdictions it may be permitted regardless of what a licence says.
+
+ADDITIONAL TERM 2, on AI tooling, is a contractual condition rather than a
+copyright one: it binds someone who accepted this licence, and it is close to
+unverifiable in practice. Treat all three as a clear statement of the copyright
+holders' position — which is what a reservation of rights is for — and not as a
+guarantee of enforcement.
+
+## Dependencies
+
+This software is an independent work that calls the public APIs of separately
+installed third-party packages. It does not incorporate, vendor, or redistribute
+their source. Those packages carry their own licences — several, including
+`lalsuite` and `pycbc`, are GPL — and installing them is the user's own act
+under their own terms. No third-party code is distributed with this package.
+
+`tdanalysis._compat.Model` is a self-contained reimplementation of the model
+base class from `mcsamplers` (https://gitlab.com/vaishakp/parameter-estimation),
+which is the work of one of the same authors and is included here under the same terms
+as the rest of this package.

tdanalysis-1.0.0/PKG-INFO

@@ -0,0 +1,174 @@
+Metadata-Version: 2.5
+Name: tdanalysis
+Version: 1.0.0
+Summary: Time-domain parameter estimation and multi-segment consistency tests for gravitational-wave data
+Project-URL: Homepage, https://github.com/vaishakp/tdanalysis
+Author: Vaishak Prasad, Viviana Cáceres
+Maintainer-email: Vaishak Prasad <vbp5193@psu.edu>
+License-Expression: LicenseRef-PolyForm-Noncommercial-1.0.0-with-AI-terms
+License-File: LICENSE
+License-File: NOTICE.md
+Keywords: consistency-test,gravitational-waves,parameter-estimation,time-domain
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: Astronomy
+Classifier: Topic :: Scientific/Engineering :: Physics
+Requires-Python: >=3.11
+Requires-Dist: bilby-pipe
+Requires-Dist: bilby<3,>=2.4
+Requires-Dist: corner
+Requires-Dist: dask
+Requires-Dist: dynesty>=2.1
+Requires-Dist: gwpy>=3.0
+Requires-Dist: h5py
+Requires-Dist: lalsuite>=7.20
+Requires-Dist: matplotlib>=3.7
+Requires-Dist: numba
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Requires-Dist: pycbc>=2.3
+Requires-Dist: scipy>=1.10
+Requires-Dist: spectools
+Requires-Dist: statsmodels
+Requires-Dist: tqdm
+Requires-Dist: waveformtools
+Provides-Extra: all
+Requires-Dist: nrhjsurrogate; extra == 'all'
+Requires-Dist: threadpoolctl; extra == 'all'
+Requires-Dist: vlconf; extra == 'all'
+Provides-Extra: gpu-legacy
+Requires-Dist: torch; extra == 'gpu-legacy'
+Provides-Extra: nr
+Requires-Dist: nrhjsurrogate; extra == 'nr'
+Provides-Extra: pipeline
+Requires-Dist: threadpoolctl; extra == 'pipeline'
+Requires-Dist: vlconf; extra == 'pipeline'
+Description-Content-Type: text/markdown
+
+# tdanalysis
+
+Time-domain parameter estimation for gravitational-wave data, and the
+inspiral-ringdown consistency test built on it.
+
+The likelihood is `-½ δᵀ C⁻¹ δ` evaluated on a chosen data segment, where `C` is
+a full noise covariance built from a detector autocorrelation function rather
+than from a diagonal-in-frequency PSD approximation. That is what makes short,
+gated or gappy segments tractable, which is the regime a ringdown analysis lives
+in.
+
+This is release **v1**. It is pure Python, it installs from PyPI with no
+compiler, and it contains the parameter estimation core and the consistency
+test. The batched GPU sampling path (v2) and the ringdown spectroscopy
+likelihoods (v3) are separate, later releases of this same package.
+
+## Install
+
+```bash
+pip install tdanalysis
+```
+
+The wheel is `py3-none-any`. Python 3.11 or newer.
+
+Optional extras:
+
+| Extra | Pulls in | For |
+|---|---|---|
+| `tdanalysis[pipeline]` | `vlconf`, `threadpoolctl` | the `tdanalysis` config-driven CLI |
+| `tdanalysis[nr]` | `nrhjsurrogate` | the NRAASur action-angle waveform backend |
+| `tdanalysis[gpu-legacy]` | `torch` | non-numba backends of the legacy matrix multiplier |
+| `tdanalysis[all]` | the first two | |
+
+Every one of these is imported inside a function at the point it is first
+needed, so the library imports and runs without any of them.
+
+`nrhjsurrogate` declares `requires-python >= 3.12`, so the `[nr]` extra can
+only be resolved on 3.12 or newer even though the package itself supports
+3.11. On an older interpreter pip fails with that as the reason, which is
+the intended behaviour: the alternative is a marker that installs nothing
+and leaves the backend quietly missing.
+
+## What is in here
+
+* **Parameter estimation.** `tdanalysis.likelihoods.single.pe.GWTransientTD` for
+  one detector, `...likelihoods.multi.pe.GWTransientTDMultiDetector` for a
+  coherent network. A bilby `Likelihood`, so it drops into a bilby sampler run.
+* **Inspiral-ringdown consistency test.**
+  `tdanalysis.likelihoods.single.consistency_test.InspiralRingdownConsistencyTD`
+  and `...likelihoods.multi.consistency_test.IMRConsistencyMultiDetector`,
+  comparing an inspiral-derived and a ringdown-derived remnant, plus
+  `tdanalysis.likelihoods.comparison.discriminators.ChiSquaredRingdown`.
+* **Quadratic forms and whiteners.** `tdanalysis.forms`, selected with
+  `qform_method`: Gohberg-Semencul (`gs`, O(N log N) via FFT), preconditioned CG
+  (`pcg`), `invLD`, `Llt` and `CD`. `gs_batched` is the numpy batched chi-square
+  that `reduced_batch` drives.
+* **Covariance and PSD handling.** `tdanalysis.covariance`, including the band
+  patching and taper conditioning the tests cover.
+* **Next-generation detectors.** `tdanalysis.detectors.xg`: CE40, CE20, the
+  triangular ET (ET1/ET2/ET3) and LIGO-India, with vendored noise curves and
+  conditioned autocorrelations under `tdanalysis/data/`. Importing
+  `tdanalysis.detectors` registers them with pycbc. Provenance for every
+  vendored file is in `tdanalysis/data/*/PROVENANCE.md`.
+* **Waveform backends.** LAL and pycbc approximants, plus the NRAASur
+  action-angle surrogate (`tdanalysis.waveforms.nraasur`). All wrapped at the
+  generator level, so peak finding, tapers, padding, the segment crop and the
+  projection are the same code for every approximant. That is what makes
+  parameters estimated with different models comparable.
+* **A CLI.** `tdanalysis --help`, a config-driven run driver.
+
+Not in v1: ringdown spectroscopy likelihoods and Kerr QNM templates (v3), the
+Kokkos GPU kernels and batched sampling (v2).
+
+## Worked example
+
+`tests/test_pe_smoke.py` is the reference example as well as the release gate.
+It builds a real `GWTransientTD` with real LAL waveform generation, real
+conditioning and projection, and a real Gohberg-Semencul form from a real
+Cholesky factor, with nothing mocked. It carries no data-file dependency, so it
+runs anywhere the package installs.
+
+Read it for the constructor arguments in context. The properties it pins are
+worth knowing about:
+
+* the zero-noise truth is the maximum, and moving off it costs chi-square
+* scalar and batched evaluation give the same number, so a posterior does not
+  depend on the batch size
+* batching is row-order independent, and a one-row batch matches the scalar call
+* the likelihood survives a pickle round trip, which is what lets a bilby
+  multiprocessing pool carry it
+
+## Tests
+
+```bash
+pip install tdanalysis
+python -m pytest tests/ -q
+```
+
+Five files ship: the `GWTransientTD` gate above, plus PSD band patching, taper
+conditioning, whitening backend agreement and the XG autocorrelations.
+
+## Licence
+
+**PolyForm Noncommercial License 1.0.0, with two additional terms.** See
+`LICENSE` for the text and `NOTICE.md` for what it means in practice, and for
+the contributors to this release.
+
+This is deliberately not an open-source licence, because it restricts the field
+of use.
+
+* **Noncommercial use is free and needs no permission** — research, teaching,
+  and publishing results obtained with it. The licence grants that outright,
+  including where the funding is industrial.
+* **Commercial use needs a written licence**, which may carry a fee or a
+  royalty: vbp5193@psu.edu.
+* **Training a model on this software is not permitted**, commercial or not.
+  Building a surrogate or emulator of the *waveforms and posteriors it produces*
+  is ordinary research use and is fine; ingesting the source and documentation
+  into a general-purpose code or language model is not.
+* **AI tools may be used on this software only with training and retention
+  off.** AI assistance as such is not prohibited; the configuration that would
+  feed this code into a training corpus is.
+
+Because those additional terms modify PolyForm, the metadata declares
+`LicenseRef-PolyForm-Noncommercial-1.0.0-with-AI-terms` rather than the bare
+PolyForm identifier.

tdanalysis-1.0.0/README.md

@@ -0,0 +1,126 @@
+# tdanalysis
+
+Time-domain parameter estimation for gravitational-wave data, and the
+inspiral-ringdown consistency test built on it.
+
+The likelihood is `-½ δᵀ C⁻¹ δ` evaluated on a chosen data segment, where `C` is
+a full noise covariance built from a detector autocorrelation function rather
+than from a diagonal-in-frequency PSD approximation. That is what makes short,
+gated or gappy segments tractable, which is the regime a ringdown analysis lives
+in.
+
+This is release **v1**. It is pure Python, it installs from PyPI with no
+compiler, and it contains the parameter estimation core and the consistency
+test. The batched GPU sampling path (v2) and the ringdown spectroscopy
+likelihoods (v3) are separate, later releases of this same package.
+
+## Install
+
+```bash
+pip install tdanalysis
+```
+
+The wheel is `py3-none-any`. Python 3.11 or newer.
+
+Optional extras:
+
+| Extra | Pulls in | For |
+|---|---|---|
+| `tdanalysis[pipeline]` | `vlconf`, `threadpoolctl` | the `tdanalysis` config-driven CLI |
+| `tdanalysis[nr]` | `nrhjsurrogate` | the NRAASur action-angle waveform backend |
+| `tdanalysis[gpu-legacy]` | `torch` | non-numba backends of the legacy matrix multiplier |
+| `tdanalysis[all]` | the first two | |
+
+Every one of these is imported inside a function at the point it is first
+needed, so the library imports and runs without any of them.
+
+`nrhjsurrogate` declares `requires-python >= 3.12`, so the `[nr]` extra can
+only be resolved on 3.12 or newer even though the package itself supports
+3.11. On an older interpreter pip fails with that as the reason, which is
+the intended behaviour: the alternative is a marker that installs nothing
+and leaves the backend quietly missing.
+
+## What is in here
+
+* **Parameter estimation.** `tdanalysis.likelihoods.single.pe.GWTransientTD` for
+  one detector, `...likelihoods.multi.pe.GWTransientTDMultiDetector` for a
+  coherent network. A bilby `Likelihood`, so it drops into a bilby sampler run.
+* **Inspiral-ringdown consistency test.**
+  `tdanalysis.likelihoods.single.consistency_test.InspiralRingdownConsistencyTD`
+  and `...likelihoods.multi.consistency_test.IMRConsistencyMultiDetector`,
+  comparing an inspiral-derived and a ringdown-derived remnant, plus
+  `tdanalysis.likelihoods.comparison.discriminators.ChiSquaredRingdown`.
+* **Quadratic forms and whiteners.** `tdanalysis.forms`, selected with
+  `qform_method`: Gohberg-Semencul (`gs`, O(N log N) via FFT), preconditioned CG
+  (`pcg`), `invLD`, `Llt` and `CD`. `gs_batched` is the numpy batched chi-square
+  that `reduced_batch` drives.
+* **Covariance and PSD handling.** `tdanalysis.covariance`, including the band
+  patching and taper conditioning the tests cover.
+* **Next-generation detectors.** `tdanalysis.detectors.xg`: CE40, CE20, the
+  triangular ET (ET1/ET2/ET3) and LIGO-India, with vendored noise curves and
+  conditioned autocorrelations under `tdanalysis/data/`. Importing
+  `tdanalysis.detectors` registers them with pycbc. Provenance for every
+  vendored file is in `tdanalysis/data/*/PROVENANCE.md`.
+* **Waveform backends.** LAL and pycbc approximants, plus the NRAASur
+  action-angle surrogate (`tdanalysis.waveforms.nraasur`). All wrapped at the
+  generator level, so peak finding, tapers, padding, the segment crop and the
+  projection are the same code for every approximant. That is what makes
+  parameters estimated with different models comparable.
+* **A CLI.** `tdanalysis --help`, a config-driven run driver.
+
+Not in v1: ringdown spectroscopy likelihoods and Kerr QNM templates (v3), the
+Kokkos GPU kernels and batched sampling (v2).
+
+## Worked example
+
+`tests/test_pe_smoke.py` is the reference example as well as the release gate.
+It builds a real `GWTransientTD` with real LAL waveform generation, real
+conditioning and projection, and a real Gohberg-Semencul form from a real
+Cholesky factor, with nothing mocked. It carries no data-file dependency, so it
+runs anywhere the package installs.
+
+Read it for the constructor arguments in context. The properties it pins are
+worth knowing about:
+
+* the zero-noise truth is the maximum, and moving off it costs chi-square
+* scalar and batched evaluation give the same number, so a posterior does not
+  depend on the batch size
+* batching is row-order independent, and a one-row batch matches the scalar call
+* the likelihood survives a pickle round trip, which is what lets a bilby
+  multiprocessing pool carry it
+
+## Tests
+
+```bash
+pip install tdanalysis
+python -m pytest tests/ -q
+```
+
+Five files ship: the `GWTransientTD` gate above, plus PSD band patching, taper
+conditioning, whitening backend agreement and the XG autocorrelations.
+
+## Licence
+
+**PolyForm Noncommercial License 1.0.0, with two additional terms.** See
+`LICENSE` for the text and `NOTICE.md` for what it means in practice, and for
+the contributors to this release.
+
+This is deliberately not an open-source licence, because it restricts the field
+of use.
+
+* **Noncommercial use is free and needs no permission** — research, teaching,
+  and publishing results obtained with it. The licence grants that outright,
+  including where the funding is industrial.
+* **Commercial use needs a written licence**, which may carry a fee or a
+  royalty: vbp5193@psu.edu.
+* **Training a model on this software is not permitted**, commercial or not.
+  Building a surrogate or emulator of the *waveforms and posteriors it produces*
+  is ordinary research use and is fine; ingesting the source and documentation
+  into a general-purpose code or language model is not.
+* **AI tools may be used on this software only with training and retention
+  off.** AI assistance as such is not prohibited; the configuration that would
+  feed this code into a training corpus is.
+
+Because those additional terms modify PolyForm, the metadata declares
+`LicenseRef-PolyForm-Noncommercial-1.0.0-with-AI-terms` rather than the bare
+PolyForm identifier.