derivkit 1.0.2__tar.gz → 1.2.0__tar.gz

{derivkit-1.0.2 → derivkit-1.2.0}/.github/workflows/ci.yml

@@ -26,7 +26,7 @@ jobs:
           python-version: ${{ matrix.python-version }}
 
       - name: Install dependencies
-        run: uv sync --extra test
+        run: uv sync --group test
 
       - name: Run tests
         run: uv run tox -m test
@@ -46,7 +46,7 @@ jobs:
           version: "0.8.22" # Fix the version of uv
 
       - name: Install dependencies
-        run: uv sync --extra lint
+        run: uv sync --group lint
 
       - name: Run ruff
         run: uv run tox -e lint -- --output-format github

{derivkit-1.0.2 → derivkit-1.2.0}/.github/workflows/docs.yml

@@ -13,7 +13,9 @@ jobs:
     permissions:
       contents: write
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v6
+        with:
+          fetch-depth: 0
 
       - uses: actions/setup-python@v3
         with:
@@ -21,7 +23,7 @@ jobs:
 
       - name: Install dependencies
         run: |
-          pip install -e .
+          pip install --group dev -e .
 
       - name: Sphinx build
         run: |

derivkit-1.2.0/CHANGELOG.md

@@ -0,0 +1,36 @@
+# Changelog
+
+## v1.2.0
+
+### Added
+
+* Added thread-safety options to `ForecastKit` and `CalculusKit`.
+* Added a multiprocessing support to `parallel_execute()`, which avoids the Python GIL.
+* Added input-level caching of function evaluations.
+
+### Changed
+
+* Exposed the $X-Y$ Gaussian Fisher matrix through `ForecastKit.xy_fisher()`.
+* Improved the general documentation.
+
+
+## v1.1.0
+
+### Added
+
+* Added caching of function and derivative values for forecasting.
+
+
+## v1.0.2
+
+### Fixed
+
+* Fixed how `dk_kwargs` are passed to `ForecastKit`.
+
+
+## v1.0.1
+
+### Fixed
+
+* Fixed an issue where the Hessian routine would repeatedly evaluate the input function if the function is vector-valued.
+* Fixed an issue where the Hyper-Hessian routine would repeatedly evaluate the input function if the function is vector-valued.

derivkit-1.2.0/CITATION.cff

@@ -0,0 +1,40 @@
+cff-version: 1.2.0
+message: "If you use DerivKit, please cite it as below."
+title: "DerivKit"
+type: software
+
+authors:
+  - given-names: "Nikolina"
+    family-names: "Šarčević"
+    email: "nikolina.sarcevic@gmail.com"
+    orcid: "0000-0001-7301-6415"
+  - given-names: "Matthijs"
+    family-names: "van der Wild"
+    email: "matthijs@vanderwild.com"
+    orcid: "0000-0002-3949-3063"
+  - given-names: "Cynthia"
+    family-names: "Trendafilova"
+    email: "cyntrendafilova@gmail.com"
+    orcid: "0000-0001-5500-4058"
+
+license: "MIT"
+repository-code: "https://github.com/derivkit/derivkit"
+
+preferred-citation:
+  type: article
+  title: "DerivKit: stable numerical derivatives bridging Fisher forecasts and MCMC"
+  authors:
+    - given-names: "Nikolina"
+      family-names: "Šarčević"
+      orcid: "https://orcid.org/0000-0001-7301-6415"
+    - given-names: "Matthijs"
+      family-names: "van der Wild"
+      orcid: "https://orcid.org/0000-0002-3949-3063"
+    - given-names: "Cynthia"
+      family-names: "Trendafilova"
+      orcid: "https://orcid.org/0000-0001-5500-4058"
+  year: 2026
+  url: "https://arxiv.org/abs/2602.08078"
+  identifiers:
+    - type: other
+      value: "arXiv:2602.08078"

{derivkit-1.0.2 → derivkit-1.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: derivkit
-Version: 1.0.2
+Version: 1.2.0
 Summary: A robust toolkit for stable numerical derivatives
 Author-email: Nikolina Šarčević <nikolina.sarcevic@gmail.com>, Matthijs van der Wild <matthijs@vanderwild.com>, Cynthia Trendafilova <cyntrendafilova@gmail.com>
 License-Expression: MIT
@@ -16,19 +16,8 @@ Requires-Dist: numdifftools
 Requires-Dist: numpy
 Requires-Dist: multiprocess
 Requires-Dist: scipy
-Requires-Dist: tox
 Requires-Dist: getdist
 Requires-Dist: emcee
-Provides-Extra: test
-Requires-Dist: pytest; extra == "test"
-Provides-Extra: docs
-Requires-Dist: sphinx<9; extra == "docs"
-Requires-Dist: sphinx-design; extra == "docs"
-Requires-Dist: sphinx-multiversion; extra == "docs"
-Requires-Dist: furo; extra == "docs"
-Requires-Dist: sphinx-copybutton; extra == "docs"
-Provides-Extra: lint
-Requires-Dist: ruff; extra == "lint"
 Provides-Extra: jax
 Requires-Dist: jax>=0.4; extra == "jax"
 Requires-Dist: jaxlib>=0.4; extra == "jax"
@@ -36,9 +25,10 @@ Dynamic: license-file
 
 # derivkit
 
-[![CI](https://github.com/derivkit/derivkit/actions/workflows/ci.yml/badge.svg)](https://github.com/derivkit/derivkit/actions/workflows/ci.yml)
-[![Deploy Sphinx Documentation](https://github.com/derivkit/derivkit/actions/workflows/docs.yml/badge.svg)](https://github.com/derivkit/derivkit/actions/workflows/docs.yml)
-[![License](https://img.shields.io/github/license/derivkit/derivkit)](LICENSE)
+[![CI](https://img.shields.io/github/actions/workflow/status/derivkit/derivkit/ci.yml?branch=main&label=CI&color=3b9ab2&style=flat-square)](https://github.com/derivkit/derivkit/actions/workflows/ci.yml)
+[![Docs](https://img.shields.io/github/actions/workflow/status/derivkit/derivkit/docs.yml?branch=main&label=docs&color=3b9ab2&style=flat-square)](https://github.com/derivkit/derivkit/actions/workflows/docs.yml)
+[![License](https://img.shields.io/github/license/derivkit/derivkit?color=3b9ab2&style=flat-square)](LICENSE)
+[![PyPI](https://img.shields.io/pypi/v/derivkit?color=3b9ab2&style=flat-square)](https://pypi.org/project/derivkit/)
 
 
 **DerivKit** is a robust Python toolkit for stable numerical derivatives, built for scientific computing, cosmology, and any domain requiring accurate gradients or higher-order expansions.
@@ -46,5 +36,22 @@ Dynamic: license-file
 Detailed documentation, examples, and API reference can be found at the [derivkit documentation](https://docs.derivkit.org).
 
 
+## Citation
+
+If you use **DerivKit** in your research, please cite the associated paper:
+
+```bibtex
+@misc{sarcevic2026derivkitstablenumericalderivatives,
+  title        = {DerivKit: stable numerical derivatives bridging Fisher forecasts and MCMC},
+  author       = {Šarčević, Nikolina and van der Wild, Matthijs and Trendafilova, Cynthia},
+  year         = {2026},
+  eprint       = {2602.08078},
+  archivePrefix= {arXiv},
+  primaryClass = {astro-ph.IM},
+  url          = {https://arxiv.org/abs/2602.08078}
+}
+```
+
+
 ## License
 MIT License © 2025 Niko Šarčević, Matthijs van der Wild, Cynthia Trendafilova and contributors.

derivkit-1.2.0/README.md

@@ -0,0 +1,32 @@
+# derivkit
+
+[![CI](https://img.shields.io/github/actions/workflow/status/derivkit/derivkit/ci.yml?branch=main&label=CI&color=3b9ab2&style=flat-square)](https://github.com/derivkit/derivkit/actions/workflows/ci.yml)
+[![Docs](https://img.shields.io/github/actions/workflow/status/derivkit/derivkit/docs.yml?branch=main&label=docs&color=3b9ab2&style=flat-square)](https://github.com/derivkit/derivkit/actions/workflows/docs.yml)
+[![License](https://img.shields.io/github/license/derivkit/derivkit?color=3b9ab2&style=flat-square)](LICENSE)
+[![PyPI](https://img.shields.io/pypi/v/derivkit?color=3b9ab2&style=flat-square)](https://pypi.org/project/derivkit/)
+
+
+**DerivKit** is a robust Python toolkit for stable numerical derivatives, built for scientific computing, cosmology, and any domain requiring accurate gradients or higher-order expansions.
+
+Detailed documentation, examples, and API reference can be found at the [derivkit documentation](https://docs.derivkit.org).
+
+
+## Citation
+
+If you use **DerivKit** in your research, please cite the associated paper:
+
+```bibtex
+@misc{sarcevic2026derivkitstablenumericalderivatives,
+  title        = {DerivKit: stable numerical derivatives bridging Fisher forecasts and MCMC},
+  author       = {Šarčević, Nikolina and van der Wild, Matthijs and Trendafilova, Cynthia},
+  year         = {2026},
+  eprint       = {2602.08078},
+  archivePrefix= {arXiv},
+  primaryClass = {astro-ph.IM},
+  url          = {https://arxiv.org/abs/2602.08078}
+}
+```
+
+
+## License
+MIT License © 2025 Niko Šarčević, Matthijs van der Wild, Cynthia Trendafilova and contributors.

{derivkit-1.0.2 → derivkit-1.2.0}/docs/_static/derivkit.css

@@ -193,4 +193,39 @@ div.plot-directive img,
 div.figure img {
   max-width: 520px;
   height: auto;
-}
+}
+
+/* ==========================================================================
+   DerivKit — Furo sidebar typography
+   ========================================================================== */
+
+/* Slightly larger sidebar base text */
+.sidebar-container {
+  font-size: 1.02rem !important;
+}
+
+/* Main navigation links (Furo) */
+.sidebar-tree a {
+  font-size: 1.08rem !important;
+  line-height: 1.4 !important;
+}
+
+/* Section headers / captions inside the sidebar */
+.sidebar-tree .caption-text {
+  font-size: 1.05rem !important;
+  font-weight: 600 !important;
+}
+
+/* Search box text */
+.sidebar-search input {
+  font-size: 1.02rem !important;
+}
+
+/* Optional: tiny spacing improvement */
+.sidebar-tree li {
+  margin-bottom: 0.15rem !important;
+}
+
+.sidebar-tree a {
+  padding: 0.15rem 0 !important;
+}

{derivkit-1.0.2 → derivkit-1.2.0}/docs/_templates/sidebar/variant-selector.html

@@ -6,11 +6,13 @@
     <li class="toctree-l1"><a class="reference" href="{{ item.url }}">{{ item.name }}</a></li>
     {%- endfor %}
   </ul>
-  <p class="caption" role="heading"><span class="caption-text">{{ _('Tags') }}</span></p>
-  <ul>
-    {%- for item in versions.tags %}
-    <li class="toctree-l1"><a class="reference" href="{{ item.url }}">{{ item.name }}</a></li>
-    {%- endfor %}
-  </ul>
+  <details>
+    <summary class="caption" role="heading"><span class="caption-text">{{ _('Versions') }}</span></summary>
+    <ul>
+      {%- for item in versions.tags | reverse %}
+      <li class="toctree-l1"><a class="reference" href="{{ item.url }}">{{ item.name }}</a></li>
+      {%- endfor %}
+    </ul>
+  </details>
 </div>
 {% endif %}

{derivkit-1.0.2 → derivkit-1.2.0}/docs/about/kits/forecast_kit.rst

@@ -197,6 +197,8 @@ Examples
 A worked example is provided in :doc:`../../examples/forecasting/fisher_xy`.
 
 
+.. _forecastkit-fisher-bias:
+
 Fisher Bias
 -----------
 
@@ -228,12 +230,34 @@ ForecastKit returns:
 Interpretation
 ^^^^^^^^^^^^^^
 
-Fisher bias estimates how small systematic errors in the observables translate
-into shifts in best-fit parameters.
+Fisher bias characterizes the impact of model incompleteness rather than explicit
+parameter perturbations. The difference vector :math:`\Delta\nu` represents a
+systematic mismatch between the true data-generating model and the model assumed
+in analysis, evaluated at the same fiducial parameters. The resulting shift
+:math:`\Delta\theta` estimates how the inferred best-fit parameters would
+move in order to absorb this mismatch if the systematic were neglected.
+Fisher bias should therefore be interpreted as a diagnostic of whether a
+given systematic can be safely ignored within a local Gaussian approximation.
 
 .. image:: ../../assets/plots/fisher_bias_demo_1and2sigma.png
    :width: 60%
 
+In practice, the two data vectors entering :math:`\Delta\nu` are obtained
+by evaluating two models at the same fiducial parameters: a baseline model used
+for inference and a more complete model that includes additional physical or
+instrumental systematics. Fisher bias then estimates how the inferred parameters
+would shift if data generated by the more complete model were analyzed using the
+baseline model alone. In this sense, Fisher bias quantifies the parameter impact
+of neglecting specific systematic effects, rather than correcting for them.
+
+The difference vector :math:`\Delta\nu_i` can be defined with either sign convention, *e.g.*
+:math:`\pm\Delta\nu_i = \nu_i^{\mathrm{biased}} - \nu_i^{\mathrm{unbiased}}`.
+The chosen convention fixes the interpretation of the resulting
+parameter shift :math:`\Delta\theta`: changing the sign of :math:`\Delta\nu`
+simply reverses the direction of the inferred bias and must be interpreted
+accordingly.
+
+
 Examples
 ^^^^^^^^
 

derivkit-1.2.0/docs/api/derivkit.utils.caching.rst

@@ -0,0 +1,7 @@
+derivkit.utils.caching module
+=============================
+
+.. automodule:: derivkit.utils.caching
+   :members:
+   :show-inheritance:
+   :undoc-members:

{derivkit-1.0.2 → derivkit-1.2.0}/docs/api/derivkit.utils.rst

@@ -7,12 +7,14 @@ Submodules
 .. toctree::
    :maxdepth: 2
 
+   derivkit.utils.caching
    derivkit.utils.concurrency
    derivkit.utils.extrapolation
    derivkit.utils.linalg
    derivkit.utils.logger
    derivkit.utils.numerics
    derivkit.utils.sandbox
+   derivkit.utils.thread_safety
    derivkit.utils.types
    derivkit.utils.validate
 

derivkit-1.2.0/docs/api/derivkit.utils.thread_safety.rst

@@ -0,0 +1,7 @@
+derivkit.utils.thread\_safety module
+====================================
+
+.. automodule:: derivkit.utils.thread_safety
+   :members:
+   :show-inheritance:
+   :undoc-members:

{derivkit-1.0.2 → derivkit-1.2.0}/docs/citation.rst

@@ -14,11 +14,14 @@ How to cite DerivKit
 
 .. code-block:: bibtex
 
-   @software{Sarcevic_DerivKit,
-     author  = {Šarčević, Nikolina and van der Wild, Matthijs and Trendafilova, Cynthia},
-     title   = {{DerivKit}},
-     license = {MIT},
-     url     = {https://github.com/derivkit/derivkit}
+   @misc{sarcevic2026derivkitstablenumericalderivatives,
+     title         = {DerivKit: stable numerical derivatives bridging Fisher forecasts and MCMC},
+     author        = {Šarčević, Nikolina and van der Wild, Matthijs and Trendafilova, Cynthia},
+     year          = {2026},
+     eprint        = {2602.08078},
+     archivePrefix = {arXiv},
+     primaryClass  = {astro-ph.IM},
+     url           = {https://arxiv.org/abs/2602.08078}
    }
 
 

{derivkit-1.0.2 → derivkit-1.2.0}/docs/contributing.rst

@@ -50,7 +50,9 @@ Running ``derivkit`` workflows
 ******************************
 
 ``derivkit`` uses `tox <https://tox.wiki>`__ to run its workflows.
-It is installed along with  ``derivkit`` itself.
+It can be installed along with ``derivkit`` by adding the ``dev`` group::
+
+  pip install --group dev derivkit
 
 All workflows can be run consecutively by simply calling tox from the project root directory::
 

{derivkit-1.0.2 → derivkit-1.2.0}/docs/examples/calculus/gradient.rst

@@ -38,7 +38,7 @@ Basic usage
 .. doctest:: gradient_basic
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[1] ** 2
@@ -60,7 +60,7 @@ Finite differences (Ridders) via ``dk_kwargs``
 .. doctest:: gradient_finite_ridders
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[1] ** 2
@@ -83,7 +83,7 @@ Adaptive backend via ``dk_kwargs``
 .. doctest:: gradient_adaptive
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[1] ** 2
@@ -108,7 +108,7 @@ The number of parallel processes can be tuned with the ``n_workers`` parameter.
 .. doctest:: gradient_parallel
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def f(theta):
    ...     return np.sin(theta[0]) + theta[1] ** 2 + np.cos(theta[2])

{derivkit-1.0.2 → derivkit-1.2.0}/docs/examples/calculus/hessian.rst

@@ -43,7 +43,7 @@ Basic usage (scalar-valued function)
 .. doctest:: hessian_basic
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[0] * theta[1] + theta[1] ** 2
@@ -75,7 +75,7 @@ DerivKit provides a fast helper for this case.
 .. doctest:: hessian_diag
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[0] * theta[1] + theta[1] ** 2
@@ -100,7 +100,7 @@ The result is reshaped back to ``(*out_shape, p, p)``.
 .. doctest:: hessian_tensor
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a tensor-valued function
    >>> def func(theta):
    ...     return np.array([
@@ -134,7 +134,7 @@ Finite differences (Ridders) via ``dk_kwargs``
 .. doctest:: hessian_finite_ridders
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function
    >>> def func(theta):
    ...     return np.sin(theta[0]) + theta[0] * theta[1] + theta[1] ** 2

{derivkit-1.0.2 → derivkit-1.2.0}/docs/examples/calculus/hyperhessian.rst

@@ -39,13 +39,14 @@ Basic usage
 .. doctest:: hyper_hessian_basic
 
    >>> import numpy as np
-   >>> from derivkit.calculus.hyper_hessian import build_hyper_hessian
+   >>> from derivkit import CalculusKit
    >>> # Define a scalar-valued function with non-zero third derivatives.
    >>> def func(theta):
    ...     return theta[0] ** 3 + 2.0 * theta[0] * theta[1] + theta[1] ** 2
    >>> theta0 = np.array([0.5, 2.0])
-   >>> hess3 = build_hyper_hessian(func, theta0)
-   >>> print(hess3.shape)
+   >>> ckit = CalculusKit(func, x0=theta0)
+   >>> hess3 = ckit.hyper_hessian()
+   >>> hess3.shape
    (2, 2, 2)
    >>> # Analytic reference:
    >>> # d^3 f / dtheta0^3 = 6, and all other third partials are zero.
@@ -55,6 +56,7 @@ Basic usage
    True
 
 
+
 Symmetry
 --------
 
@@ -64,13 +66,12 @@ its parameter indices:
 .. doctest:: hyper_hessian_symmetry
 
    >>> import numpy as np
-   >>> from derivkit.calculus.hyper_hessian import build_hyper_hessian
-   >>>
+   >>> from derivkit import CalculusKit
    >>> def func(theta):
    ...     return theta[0] ** 3 + 2.0 * theta[0] * theta[1] + theta[1] ** 2
-   >>>
    >>> theta0 = np.array([0.5, 2.0])
-   >>> hess3 = build_hyper_hessian(func, theta0)
+   >>> ckit = CalculusKit(func, x0=theta0)
+   >>> hess3 = ckit.hyper_hessian()
    >>> # Check symmetry for a representative index triplet.
    >>> bool(hess3[0, 1, 0] == hess3[1, 0, 0] == hess3[0, 0, 1])
    True
@@ -85,7 +86,7 @@ for each output component and afterwards reshaped back to ``(*out_shape, p, p, p
 .. doctest:: hyper_hessian_tensor
 
    >>> import numpy as np
-   >>> from derivkit.calculus.hyper_hessian import build_hyper_hessian
+   >>> from derivkit import CalculusKit
    >>> # Two-output function.
    >>> # f0(theta) = theta0^3
    >>> # f1(theta) = theta0*theta1 + theta1^2
@@ -93,10 +94,11 @@ for each output component and afterwards reshaped back to ``(*out_shape, p, p, p
    ...     return np.array([
    ...         theta[0] ** 3,
    ...         theta[0] * theta[1] + theta[1] ** 2,
-   ...     ])
+   ...     ], dtype=float)
    >>> theta0 = np.array([0.5, 2.0])
-   >>> hess3 = build_hyper_hessian(func, theta0)
-   >>> print(hess3.shape)
+   >>> ckit = CalculusKit(func, x0=theta0)
+   >>> hess3 = ckit.hyper_hessian()
+   >>> hess3.shape
    (2, 2, 2, 2)
    >>>
    >>> # Component 0: only d^3/dtheta0^3 = 6 is non-zero.
@@ -118,13 +120,12 @@ options via ``**dk_kwargs``.
 .. doctest:: hyper_hessian_finite_ridders
 
    >>> import numpy as np
-   >>> from derivkit.calculus.hyper_hessian import build_hyper_hessian
+   >>> from derivkit import CalculusKit
    >>> def func(theta):
    ...     return theta[0] ** 3 + 2.0 * theta[0] * theta[1] + theta[1] ** 2
    >>> theta0 = np.array([0.5, 2.0])
-   >>> hess3 = build_hyper_hessian(
-   ...     func,
-   ...     theta0,
+   >>> ckit = CalculusKit(func, x0=theta0)
+   >>> hess3 = ckit.hyper_hessian(
    ...     method="finite",
    ...     n_workers=4,
    ...     extrapolation="ridders",

{derivkit-1.0.2 → derivkit-1.2.0}/docs/examples/calculus/jacobian.rst

@@ -42,7 +42,7 @@ Basic usage
 .. doctest:: jacobian_basic
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a vector-valued function
    >>> def func(theta):
    ...     return np.array([
@@ -74,7 +74,7 @@ Finite differences (Ridders) via ``dk_kwargs``
 .. doctest:: jacobian_finite_ridders
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a vector-valued function
    >>> def func(theta):
    ...     return np.array([np.sin(theta[0]) + theta[1], theta[0] * theta[1]])
@@ -99,7 +99,7 @@ Adaptive backend via ``dk_kwargs``
 .. doctest:: jacobian_adaptive
 
    >>> import numpy as np
-   >>> from derivkit.calculus_kit import CalculusKit
+   >>> from derivkit import CalculusKit
    >>> # Define a vector-valued function
    >>> def func(theta):
    ...     return np.array([np.sin(theta[0]) + theta[1], theta[0] * theta[1]])

{derivkit-1.0.2 → derivkit-1.2.0}/docs/examples/derivatives/adaptive_fit.rst

@@ -32,7 +32,7 @@ By default, an adaptive Chebyshev grid is constructed automatically around
 .. doctest:: adaptive_basic
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> # Initialize DerivativeKit with the target function and expansion point
    >>> dk = DerivativeKit(function=np.sin, x0=0.7)
    >>> # Compute the first derivative (adaptive backend is the default)
@@ -60,7 +60,7 @@ Accepted forms for ``spacing`` are:
 .. doctest:: adaptive_sampling_control
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> # Initialize DerivativeKit with the target function and expansion point
    >>> dk = DerivativeKit(function=np.sin, x0=0.7)
    >>> # Compute the first derivative with custom sampling
@@ -84,7 +84,7 @@ The adaptive grid is clipped or transformed as needed to stay inside bounds.
 .. doctest:: adaptive_domain
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> # Define a function and an expansion point
    >>> f = np.log
    >>> x0 = 0.05
@@ -115,7 +115,7 @@ If zero is missing from an offsets grid, it is inserted automatically.
 .. doctest:: adaptive_custom_grid
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> dk = DerivativeKit(function=np.sin, x0=0.7)
    >>> offsets = np.array([-2e-2, -1e-2, 1e-2, 2e-2])  # no 0 (inserted automatically)
    >>> deriv = dk.differentiate(method="adaptive", order=1, grid=("offsets", offsets))
@@ -135,7 +135,7 @@ stability.
 .. doctest:: adaptive_ridge
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> dk = DerivativeKit(function=np.sin, x0=0.7)
    >>> deriv = dk.differentiate(
    ...    method="adaptive",
@@ -159,7 +159,7 @@ You can request an error proxy and detailed diagnostics from the polynomial fit.
 .. doctest:: adaptive_diagnostics
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> dk = DerivativeKit(function=np.sin, x0=0.7)
    >>> deriv, err, diag = dk.differentiate(
    ...    method="adaptive",
@@ -186,7 +186,7 @@ function evaluations are noisy.
 .. doctest:: adaptive_noisy
 
    >>> import numpy as np
-   >>> from derivkit.derivative_kit import DerivativeKit
+   >>> from derivkit import DerivativeKit
    >>> # Deterministic "noise": sampled once on a grid and interpolated (stable across calls)
    >>> rng = np.random.default_rng(42)
    >>> x_noise = np.linspace(0.0, 2.0, 2049)