diff --git a/examples/example_empirical_distribution.ipynb b/examples/example_empirical_distribution.ipynb new file mode 100644 index 0000000..a635261 --- /dev/null +++ b/examples/example_empirical_distribution.ipynb @@ -0,0 +1,381 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# EmpiricalDistribution\n", + "\n", + "A distribution built from observed data using a density estimator (Gaussian KDE by default). \n", + "The PDF is provided analytically via the estimator; CDF and PPF are derived automatically by the characteristic graph.\n", + "\n", + "**Two key features:**\n", + "- `with_method(m)` — returns an independent copy with a different method (the original is unchanged)\n", + "- `set_method(m)` — replaces the method in-place on the same object (cache and sampler are reset automatically)" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "import matplotlib.pyplot as plt\n", + "import numpy as np\n", + "\n", + "from pysatl_core.distributions.empirical import EmpiricalDistribution, ScipyGaussianKde" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 1. Data\n", + "\n", + "A mixture of two Gaussians — a bimodal sample where KDE methods behave differently." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "sample size: 1000, range: [-4.55, 6.81]\n" + ] + } + ], + "source": [ + "rng = np.random.default_rng(42)\n", + "sample = np.concatenate(\n", + " [\n", + " rng.normal(loc=-2.5, scale=0.8, size=400),\n", + " rng.normal(loc=3.0, scale=1.2, size=600),\n", + " ]\n", + ")\n", + "\n", + "x = np.linspace(sample.min() - 1.0, sample.max() + 1.0, 500)\n", + "quant = np.linspace(0.01, 0.99, 300)\n", + "\n", + "print(f\"sample size: {len(sample)}, range: [{sample.min():.2f}, {sample.max():.2f}]\")" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 2. Basic characteristics\n", + "\n", + "PDF, CDF (∫ PDF dx via the characteristic graph), PPF (PCHIP inversion of the CDF), new samples." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "distr = EmpiricalDistribution(sample, method=ScipyGaussianKde(bandwidth=\"silverman\"))\n", + "\n", + "pdf_vals = distr.calculate_characteristic(\"pdf\", x)\n", + "cdf_vals = distr.calculate_characteristic(\"cdf\", x)\n", + "ppf_vals = distr.calculate_characteristic(\"ppf\", quant)\n", + "new_samples = distr.sample(1000)\n", + "\n", + "fig, axes = plt.subplots(1, 3, figsize=(14, 4))\n", + "fig.suptitle(\"EmpiricalDistribution — basic characteristics\", fontweight=\"bold\")\n", + "\n", + "axes[0].hist(sample, bins=50, density=True, alpha=0.4, color=\"steelblue\", label=\"original\")\n", + "axes[0].hist(new_samples, bins=50, density=True, alpha=0.35, color=\"orchid\", label=\"sampled\")\n", + "axes[0].plot(x, pdf_vals, color=\"crimson\", lw=2, label=\"PDF\")\n", + "axes[0].set_title(\"PDF + new samples\")\n", + "axes[0].legend()\n", + "\n", + "axes[1].plot(x, cdf_vals, color=\"darkorange\", lw=2)\n", + "axes[1].set_title(\"CDF (∫ PDF dx)\")\n", + "axes[1].set_ylim(-0.02, 1.02)\n", + "axes[1].grid(alpha=0.3)\n", + "\n", + "axes[2].plot(quant, ppf_vals, color=\"seagreen\", lw=2)\n", + "axes[2].set_title(\"PPF (CDF⁻¹ via PCHIP)\")\n", + "axes[2].set_xlabel(\"quantile q\")\n", + "axes[2].grid(alpha=0.3)\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 3. `with_method` — side-by-side method comparison\n", + "\n", + "`with_method(m)` returns a **new** object with a different method fitted on the same sample. \n", + "The original `distr` is unchanged. Each object's strategies are independent." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "sample shared across all clones: OK\n", + "computation strategies are independent: OK\n" + ] + } + ], + "source": [ + "methods = {\n", + " \"silverman\": distr,\n", + " \"scott\": distr.with_method(ScipyGaussianKde(bandwidth=\"scott\")),\n", + " \"bw=0.15\": distr.with_method(ScipyGaussianKde(bandwidth=0.15)),\n", + " \"bw=0.70\": distr.with_method(ScipyGaussianKde(bandwidth=0.70)),\n", + "}\n", + "\n", + "assert methods[\"silverman\"] is distr\n", + "assert methods[\"scott\"] is not distr\n", + "\n", + "for _name, d in methods.items():\n", + " assert d.data is sample\n", + "print(\"sample shared across all clones: OK\")\n", + "\n", + "strats = [d.computation_strategy for d in methods.values()]\n", + "assert len({id(s) for s in strats}) == len(strats)\n", + "print(\"computation strategies are independent: OK\")" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "colors = [\"crimson\", \"steelblue\", \"seagreen\", \"darkorange\"]\n", + "styles = [\"-\", \"--\", \"-.\", \":\"]\n", + "\n", + "fig, axes = plt.subplots(1, 2, figsize=(14, 5))\n", + "fig.suptitle(\"with_method — independent copies with different methods\", fontweight=\"bold\")\n", + "\n", + "axes[0].hist(sample, bins=50, density=True, alpha=0.25, color=\"gray\", label=\"sample\")\n", + "for (name, d), c, ls in zip(methods.items(), colors, styles, strict=False):\n", + " axes[0].plot(x, d.calculate_characteristic(\"pdf\", x), color=c, ls=ls, lw=2, label=name)\n", + "axes[0].set_title(\"PDF\")\n", + "axes[0].legend()\n", + "\n", + "for (name, d), c, ls in zip(methods.items(), colors, styles, strict=False):\n", + " axes[1].plot(quant, d.calculate_characteristic(\"ppf\", quant), color=c, ls=ls, lw=2, label=name)\n", + "axes[1].set_title(\"PPF\")\n", + "axes[1].set_xlabel(\"quantile q\")\n", + "axes[1].legend()\n", + "axes[1].grid(alpha=0.3)\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 4. `set_method` — in-place method replacement\n", + "\n", + "`set_method(m)` replaces the method **on the same object**:\n", + "- `_estimator` is replaced with `m.fit(sample)`\n", + "- `EmpiricalComputationStrategy` detects the estimator change via `is`-comparison and clears the cache on the next query\n", + "- The UNURAN sampler is explicitly reset via `invalidate()` because the C-side captured the old PDF" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "id(distr) unchanged: True\n", + "estimator replaced: 4597003904 -> 4602795600\n", + "sample reused (no copy): True\n" + ] + } + ], + "source": [ + "distr_mut = EmpiricalDistribution(sample, method=ScipyGaussianKde(bandwidth=0.7))\n", + "\n", + "id_distr_before = id(distr_mut)\n", + "id_estimator_before = id(distr_mut._estimator)\n", + "\n", + "pdf_before = distr_mut.calculate_characteristic(\"pdf\", x)\n", + "cdf_before = distr_mut.calculate_characteristic(\"cdf\", x)\n", + "ppf_before = distr_mut.calculate_characteristic(\"ppf\", quant)\n", + "\n", + "distr_mut.set_method(ScipyGaussianKde(bandwidth=0.15))\n", + "\n", + "print(f\"id(distr) unchanged: {id(distr_mut) == id_distr_before}\")\n", + "print(f\"estimator replaced: {id_estimator_before} -> {id(distr_mut._estimator)}\")\n", + "print(f\"sample reused (no copy): {distr_mut.data is sample}\")\n", + "\n", + "pdf_after = distr_mut.calculate_characteristic(\"pdf\", x)\n", + "cdf_after = distr_mut.calculate_characteristic(\"cdf\", x)\n", + "ppf_after = distr_mut.calculate_characteristic(\"ppf\", quant)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "fig, axes = plt.subplots(1, 3, figsize=(15, 4))\n", + "fig.suptitle(\"set_method — one object before and after method swap\", fontweight=\"bold\")\n", + "\n", + "kw_a = {\"color\": \"crimson\", \"lw\": 2, \"label\": \"bw=0.7 (before)\"}\n", + "kw_b = {\"color\": \"darkgreen\", \"lw\": 2, \"ls\": \"--\", \"label\": \"bw=0.15 (after)\"}\n", + "\n", + "axes[0].hist(sample, bins=50, density=True, alpha=0.25, color=\"gray\")\n", + "axes[0].plot(x, pdf_before, **kw_a)\n", + "axes[0].plot(x, pdf_after, **kw_b)\n", + "axes[0].set_title(\"PDF\")\n", + "axes[0].legend()\n", + "\n", + "axes[1].plot(x, cdf_before, **kw_a)\n", + "axes[1].plot(x, cdf_after, **kw_b)\n", + "axes[1].set_title(\"CDF\")\n", + "axes[1].set_ylim(-0.02, 1.02)\n", + "axes[1].legend()\n", + "axes[1].grid(alpha=0.3)\n", + "\n", + "axes[2].plot(quant, ppf_before, **kw_a)\n", + "axes[2].plot(quant, ppf_after, **kw_b)\n", + "axes[2].set_title(\"PPF\")\n", + "axes[2].set_xlabel(\"quantile q\")\n", + "axes[2].legend()\n", + "axes[2].grid(alpha=0.3)\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 5. `set_method` and the sampler\n", + "\n", + "After `set_method` UNURAN is rebuilt on the next `sample()` call. \n", + "The distribution of new samples matches the updated PDF." + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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", + "text/plain": [ + "
" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "source": [ + "distr_s = EmpiricalDistribution(sample, method=ScipyGaussianKde(bandwidth=0.7))\n", + "samples_wide = distr_s.sample(2000)\n", + "pdf_wide = distr_s.calculate_characteristic(\"pdf\", x)\n", + "\n", + "distr_s.set_method(ScipyGaussianKde(bandwidth=0.15))\n", + "samples_tight = distr_s.sample(2000)\n", + "pdf_tight = distr_s.calculate_characteristic(\"pdf\", x)\n", + "\n", + "fig, axes = plt.subplots(1, 2, figsize=(13, 4))\n", + "fig.suptitle(\"set_method — sampler is rebuilt after method swap\", fontweight=\"bold\")\n", + "\n", + "axes[0].hist(sample, bins=60, density=True, alpha=0.3, color=\"gray\", label=\"original\")\n", + "axes[0].hist(\n", + " samples_wide, bins=60, density=True, alpha=0.4, color=\"crimson\", label=\"sample (bw=0.7)\"\n", + ")\n", + "axes[0].plot(x, pdf_wide, color=\"darkred\", lw=2)\n", + "axes[0].set_title(\"Before set_method (bw=0.7)\")\n", + "axes[0].legend()\n", + "\n", + "axes[1].hist(sample, bins=60, density=True, alpha=0.3, color=\"gray\", label=\"original\")\n", + "axes[1].hist(\n", + " samples_tight, bins=60, density=True, alpha=0.4, color=\"darkgreen\", label=\"sample (bw=0.15)\"\n", + ")\n", + "axes[1].plot(x, pdf_tight, color=\"darkgreen\", lw=2)\n", + "axes[1].set_title(\"After set_method (bw=0.15)\")\n", + "axes[1].legend()\n", + "\n", + "plt.tight_layout()\n", + "plt.show()" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3 (ipykernel)", + "language": "python", + "name": "python3" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.14.3" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/src/pysatl_core/distributions/__init__.py b/src/pysatl_core/distributions/__init__.py index 8f17fc8..dc8c6f3 100644 --- a/src/pysatl_core/distributions/__init__.py +++ b/src/pysatl_core/distributions/__init__.py @@ -20,6 +20,13 @@ FitterMethod, ) from .distribution import Distribution +from .empirical import ( + EmpiricalComputationStrategy, + EmpiricalDistribution, + EmpiricalMethod, + FittedEmpirical, + ScipyGaussianKde, +) from .registry import * from .registry import __all__ as _registry_all from .strategies import ( @@ -39,6 +46,12 @@ "EvaluatorMethod", # distribution "Distribution", + # empirical + "EmpiricalComputationStrategy", + "EmpiricalDistribution", + "EmpiricalMethod", + "FittedEmpirical", + "ScipyGaussianKde", # strategies "ComputationStrategy", "DefaultComputationStrategy", diff --git a/src/pysatl_core/distributions/empirical/__init__.py b/src/pysatl_core/distributions/empirical/__init__.py new file mode 100644 index 0000000..0e5287f --- /dev/null +++ b/src/pysatl_core/distributions/empirical/__init__.py @@ -0,0 +1,23 @@ +""" +Empirical distribution and its computation strategy. +""" + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" + +from pysatl_core.distributions.empirical.distribution import ( + EmpiricalDistribution, + EmpiricalMethod, + FittedEmpirical, + ScipyGaussianKde, +) +from pysatl_core.distributions.empirical.strategy import EmpiricalComputationStrategy + +__all__ = [ + "EmpiricalComputationStrategy", + "EmpiricalDistribution", + "EmpiricalMethod", + "FittedEmpirical", + "ScipyGaussianKde", +] diff --git a/src/pysatl_core/distributions/empirical/distribution.py b/src/pysatl_core/distributions/empirical/distribution.py new file mode 100644 index 0000000..b85d559 --- /dev/null +++ b/src/pysatl_core/distributions/empirical/distribution.py @@ -0,0 +1,280 @@ +""" +Empirical Distribution + +Wraps a density estimator (e.g. KDE) built from observed data into a +``Distribution`` that integrates with the characteristic graph. +Providing only a PDF analytical computation is enough — the graph +derives CDF and PPF automatically via numerical fitters. +""" + +from __future__ import annotations + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" + +from dataclasses import dataclass +from typing import TYPE_CHECKING, Any, Literal, Protocol, cast + +import numpy as np +from numpy.typing import NDArray + +from pysatl_core.distributions.computations.computation import AnalyticalComputation +from pysatl_core.distributions.distribution import _KEEP, Distribution +from pysatl_core.distributions.empirical.strategy import EmpiricalComputationStrategy +from pysatl_core.distributions.strategies import ComputationStrategy, SamplingStrategy +from pysatl_core.sampling.unuran.core.unuran_sampling_strategy import DefaultUnuranSamplingStrategy +from pysatl_core.types import CharacteristicName, ComputationFunc, UnivariateContinuous + +if TYPE_CHECKING: + from pysatl_core.distributions.support import Support + + +class FittedEmpirical(Protocol): + """A fitted empirical density estimator that can evaluate PDF and CDF.""" + + def pdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + """Evaluate the probability density function at points *x*.""" + ... + + def cdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + """Evaluate the cumulative distribution function at points *x*.""" + ... + + +class EmpiricalMethod(Protocol): + """Strategy for constructing a :class:`FittedEmpirical` from a data sample.""" + + def fit(self, sample: NDArray[np.float64]) -> FittedEmpirical: + """Fit the estimator to *sample* and return an evaluable estimator.""" + ... + + +@dataclass(frozen=True) +class ScipyGaussianKde: + """ + Gaussian KDE via :func:`scipy.stats.gaussian_kde`. + + Parameters + ---------- + bandwidth : float or {"scott", "silverman"}, default "scott" + Bandwidth selection method or explicit scalar value. + """ + + bandwidth: float | Literal["scott", "silverman"] = "scott" + + def fit(self, sample: NDArray[np.float64]) -> FittedEmpirical: + from scipy.stats import gaussian_kde + + return _ScipyFittedKde(gaussian_kde(sample, bw_method=self.bandwidth)) + + +class _ScipyFittedKde: + """Thin wrapper that adapts ``scipy.stats.gaussian_kde`` to ``FittedEmpirical``.""" + + def __init__(self, kde: Any) -> None: + self._kde = kde + + def pdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + scalar_input = np.ndim(x) == 0 + x_arr = np.atleast_1d(np.asarray(x, dtype=float)) + finite = np.isfinite(x_arr) + result = np.zeros_like(x_arr) + if finite.any(): + result[finite] = self._kde.pdf(x_arr[finite]) + return result[0] if scalar_input else result + + def cdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + scalar_input = np.ndim(x) == 0 + x_arr = np.atleast_1d(np.asarray(x, dtype=float)) + result = np.where(x_arr == np.inf, 1.0, np.where(x_arr == -np.inf, 0.0, np.nan)) + finite = np.isfinite(x_arr) + if finite.any(): + result[finite] = np.array( + [self._kde.integrate_box_1d(-np.inf, xi) for xi in x_arr[finite]], + dtype=float, + ) + return result[0] if scalar_input else result + + +class EmpiricalDistribution(Distribution): + """ + A continuous univariate distribution built from an empirical density + estimator (KDE by default) fitted to observed data. + + The PDF is provided analytically via the estimator; CDF and PPF + are derived automatically by the characteristic graph (numerical + integration and root-finding respectively). + + Parameters + ---------- + sample : NDArray[np.float64] + One-dimensional array of observed values used to fit the estimator. + method : EmpiricalMethod, default ScipyGaussianKde() + Strategy used to construct the empirical density estimator. + support : Support or None, default None + Explicit support for the distribution. ``None`` leaves the support + unrestricted, which is the natural choice for Gaussian KDE. + sampling_strategy : SamplingStrategy or None, default None + Overrides the default inverse-transform sampling strategy. + computation_strategy : ComputationStrategy or None, default None + Overrides the default graph-based computation strategy. + + Examples + -------- + >>> import numpy as np + >>> rng = np.random.default_rng(0) + >>> sample = rng.normal(0, 1, 500) + >>> distr = EmpiricalDistribution(sample) + >>> distr.calculate_characteristic("pdf", np.array([0.0])) + array([0.39...]) + """ + + def __init__( + self, + sample: NDArray[np.float64], + method: EmpiricalMethod = ScipyGaussianKde(), + support: Support | None = None, + sampling_strategy: SamplingStrategy | None = None, + computation_strategy: ComputationStrategy | None = None, + ) -> None: + self._sample = np.asarray(sample, dtype=float) + self._method = method + self._estimator = method.fit(self._sample) + + # TODO: DefaultUnuranSamplingStrategy fallback calls DefaultSamplingUnivariateStrategy + # which passes the full U array to graph-derived PPF (scalar-only) and will fail. + # Fix by making the fallback use a scalar loop instead of ppf(U). + super().__init__( + distribution_type=UnivariateContinuous, + analytical_computations=self._build_analytical_computations(), + support=support, + sampling_strategy=sampling_strategy or DefaultUnuranSamplingStrategy(), + computation_strategy=computation_strategy or EmpiricalComputationStrategy(), + ) + + def _pdf(self, x: NDArray[np.float64], **_options: Any) -> NDArray[np.float64]: + """Indirection: always reads the current ``_estimator``. + + Used as the analytical PDF func so swapping ``_estimator`` (via + :meth:`set_method`) takes effect without rebuilding analytical + computations or graph loop edges. + """ + return self._estimator.pdf(x) + + def _cdf(self, x: NDArray[np.float64], **_options: Any) -> NDArray[np.float64]: + """Indirection: always reads the current ``_estimator``. See :meth:`_pdf`.""" + return self._estimator.cdf(x) + + def _build_analytical_computations( + self, + ) -> dict[str, AnalyticalComputation[Any, Any]]: + """Single source of truth for the analytical PDF/CDF entries.""" + result: dict[str, AnalyticalComputation[Any, Any]] = { + CharacteristicName.PDF: AnalyticalComputation( + target=CharacteristicName.PDF, + func=cast(ComputationFunc[Any, Any], self._pdf), + ), + CharacteristicName.CDF: AnalyticalComputation( + target=CharacteristicName.CDF, + func=cast(ComputationFunc[Any, Any], self._cdf), + ), + } + return result + + @property + def data(self) -> NDArray[np.float64]: + """The original data sample used to fit this distribution.""" + return self._sample + + @property + def method(self) -> EmpiricalMethod: + """The empirical method currently configured on this distribution.""" + return self._method + + def with_method(self, method: EmpiricalMethod) -> EmpiricalDistribution: + """ + Return a clone of this distribution with a different empirical method. + + The clone refits the new method on the same sample (shared array, no copy). + Strategies are deep-copied like in any other ``with_*`` clone, so the new + instance starts with empty caches and a fresh sampler — independent of + anything the original may have memoised. + + Use this in preference to :meth:`set_method` when you want to compare + methods side-by-side or keep the original distribution intact. + """ + return self._clone_with_strategies(method=method) + + def set_method(self, method: EmpiricalMethod) -> None: + """ + Replace the empirical method in place. + + Refits ``method`` on the original sample and rebinds the underlying + estimator. The graph's analytical loops do not need to be rebuilt: + :meth:`_pdf` and :meth:`_cdf` always read the current estimator, + and :class:`EmpiricalComputationStrategy` clears its fitted-method + cache automatically when it notices the estimator object has changed. + + Sampling strategies that hold cached state (notably + :class:`DefaultUnuranSamplingStrategy`, whose generator is built once + on the PDF at the time of first sample) are reset via their + ``invalidate()`` method when present. Strategies without + ``invalidate`` are left untouched — the assumption is that they hold + no per-distribution state. + + Notes + ----- + Any external code that holds a direct reference to internal sampler + state (e.g. a value previously read from + ``distr.sampling_strategy._sampler``) keeps that reference alive and + will continue to sample from the *previous* distribution. Re-acquire + such references after calling :meth:`set_method`. + + For side-by-side comparison of methods, prefer :meth:`with_method`. + """ + self._method = method + self._estimator = method.fit(self._sample) + # Computation cache: auto-invalidated on next query via estimator-id + # tracking in EmpiricalComputationStrategy. Custom strategies that + # implement an invalidate() hook get a chance to reset, too. + getattr(self._computation_strategy, "invalidate", lambda: None)() + # Sampling cache: must be reset explicitly — UNURAN's C-side init + # captured the old PDF and cannot be patched in place. + getattr(self._sampling_strategy, "invalidate", lambda: None)() + + def _clone_with_strategies( + self, + *, + sampling_strategy: SamplingStrategy | None | object = _KEEP, + computation_strategy: ComputationStrategy | None | object = _KEEP, + method: EmpiricalMethod | object = _KEEP, + ) -> EmpiricalDistribution: + clone = object.__new__(EmpiricalDistribution) + clone._sample = self._sample + if method is _KEEP: + clone._method = self._method + clone._estimator = self._estimator + else: + new_method = cast(EmpiricalMethod, method) + clone._method = new_method + clone._estimator = new_method.fit(self._sample) + Distribution.__init__( + clone, + distribution_type=UnivariateContinuous, + analytical_computations=clone._build_analytical_computations(), + support=self.support, + sampling_strategy=self._new_sampling_strategy(sampling_strategy=sampling_strategy), + computation_strategy=self._new_computation_strategy( + computation_strategy=computation_strategy + ), + ) + return clone + + +__all__ = [ + "EmpiricalDistribution", + "EmpiricalMethod", + "FittedEmpirical", + "ScipyGaussianKde", +] diff --git a/src/pysatl_core/distributions/empirical/strategy.py b/src/pysatl_core/distributions/empirical/strategy.py new file mode 100644 index 0000000..dba44f6 --- /dev/null +++ b/src/pysatl_core/distributions/empirical/strategy.py @@ -0,0 +1,195 @@ +""" +Computation strategy specialised for :class:`EmpiricalDistribution`. + +Adds estimator-identity tracking on top of :class:`DefaultComputationStrategy`: +fitted methods derived from a previous underlying estimator (e.g. KDE fit on +the sample) are dropped automatically when the empirical method is swapped, +so the strategy never returns a CDF/PPF derived from a stale PDF. +""" + +from __future__ import annotations + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" + +from typing import TYPE_CHECKING, Any, cast + +import numpy as np +from numpy.typing import NDArray + +from pysatl_core.distributions.computations.computation import FittedComputationMethod +from pysatl_core.distributions.strategies import DefaultComputationStrategy +from pysatl_core.types import CharacteristicName, ComputationFunc, Method + +if TYPE_CHECKING: + from collections.abc import Mapping + + from pysatl_core.distributions.computations.options import StepOptions + from pysatl_core.distributions.distribution import Distribution + from pysatl_core.types import GenericCharacteristicName + + +_PPF_GRID_SIZE = 1025 +_PPF_TAIL_MARGIN_STD = 6.0 +_PPF_QUANTILE_CLIP_EPS = 1e-6 + + +class EmpiricalComputationStrategy(DefaultComputationStrategy): + """ + Computation strategy used by :class:`EmpiricalDistribution`. + + Behaves like :class:`DefaultComputationStrategy`, except the fitted-method + cache is implicitly keyed by the identity of the distribution's underlying + estimator (``distr._estimator``). Whenever the strategy notices that the + estimator has changed since the last query, it clears the cache so that + any fitted CDF/PPF/etc. previously derived from the old estimator is + rebuilt against the new one. + + This means callers do **not** need to call an explicit ``invalidate()`` + after swapping the empirical method on the distribution — the strategy + detects the swap on the next characteristic query. + + The strategy also works for distributions without an ``_estimator`` + attribute (it then degenerates to plain Default behaviour). + """ + + def __init__(self, enable_caching: bool = False) -> None: + super().__init__(enable_caching=enable_caching) + # We hold a strong reference to the last seen estimator. Two reasons: + # (1) `is`-comparison is unambiguous (id() can be recycled by GC); + # (2) the small memory cost of keeping one extra reference is + # negligible compared to the cost of silently caching a stale fit. + self._tracked_estimator: object | None = None + # Separate cache for the vectorised PPF built by this strategy. + # Keyed by id(distr) so entries expire automatically when the + # distribution object is replaced; cleared explicitly in + # _maybe_invalidate when the estimator changes. + self._ppf_cache: dict[int, FittedComputationMethod[Any, Any]] = {} + + def _maybe_invalidate(self, distr: Distribution) -> None: + """ + Drop the cache if the distribution's estimator changed since last query. + + Distributions without an ``_estimator`` attribute (or with ``None``) are + treated as "no estimator"; the cache is reset only on transitions between + distinct estimator objects, not on every query. + """ + current = getattr(distr, "_estimator", None) + if current is not self._tracked_estimator: + self._cache.clear() + self._ppf_cache.clear() + self._tracked_estimator = current + + def query_method( + self, + state: GenericCharacteristicName, + distr: Distribution, + options: StepOptions | None = None, + *, + characteristic_options: Mapping[str, Any] | None = None, + computation_defaults: Mapping[str, Any] | None = None, + ) -> Method[Any, Any]: + self._maybe_invalidate(distr) + + if state == CharacteristicName.PPF and self._can_build_vectorized_ppf(distr): + distr_id = id(distr) + cached = self._ppf_cache.get(distr_id) + if cached is not None: + return cached + fitted = self._build_vectorized_ppf( + distr, + options, + characteristic_options=characteristic_options, + computation_defaults=computation_defaults, + ) + if self._enable_caching: + self._ppf_cache[distr_id] = fitted + return fitted + + return super().query_method( + state, + distr, + options, + characteristic_options=characteristic_options, + computation_defaults=computation_defaults, + ) + + @staticmethod + def _can_build_vectorized_ppf(distr: Distribution) -> bool: + """The vectorised PPF needs a 1-D sample and a way to evaluate the CDF.""" + sample = getattr(distr, "_sample", None) + if sample is None: + return False + sample_arr = np.asarray(sample, dtype=float) + if sample_arr.ndim != 1 or sample_arr.size == 0: + return False + return CharacteristicName.CDF in distr.analytical_computations or ( + CharacteristicName.PDF in distr.analytical_computations + ) + + def _build_vectorized_ppf( + self, + distr: Distribution, + options: StepOptions | None = None, + *, + characteristic_options: Mapping[str, Any] | None = None, + computation_defaults: Mapping[str, Any] | None = None, + ) -> FittedComputationMethod[Any, Any]: + """ + Tabulate the CDF on a sample-aware grid and invert via PCHIP. + + Cheaper than per-point root-finding when the caller passes an array of + quantiles, and naturally vectorised. PCHIP preserves the monotonicity + of the input CDF so the inverse is well-defined as long as the + tabulated CDF is strictly increasing, which we enforce by clipping + numerical noise. + """ + from scipy.interpolate import PchipInterpolator + + sample: NDArray[np.float64] = np.asarray(distr._sample, dtype=np.float64) # type: ignore[attr-defined] + sample_min = float(sample.min()) + sample_max = float(sample.max()) + # `std=0` only happens for a degenerate constant sample; in that case + # the KDE collapses to a point mass and PPF is the constant itself. + std = float(sample.std(ddof=0)) or 1.0 + margin = _PPF_TAIL_MARGIN_STD * std + x_grid = np.linspace(sample_min - margin, sample_max + margin, _PPF_GRID_SIZE) + + cdf_method = super().query_method( + CharacteristicName.CDF, + distr, + options, + characteristic_options=characteristic_options, + computation_defaults=computation_defaults, + ) + cdf_values = np.asarray(cdf_method(x_grid), dtype=float) + + cdf_values = np.maximum.accumulate(cdf_values) + cdf_values = np.clip(cdf_values, _PPF_QUANTILE_CLIP_EPS, 1.0 - _PPF_QUANTILE_CLIP_EPS) + eps = np.finfo(float).eps + for i in range(1, cdf_values.size): + if cdf_values[i] <= cdf_values[i - 1]: + cdf_values[i] = cdf_values[i - 1] + eps + + interpolator = PchipInterpolator(cdf_values, x_grid, extrapolate=False) + cdf_lo, cdf_hi = float(cdf_values[0]), float(cdf_values[-1]) + x_lo, x_hi = float(x_grid[0]), float(x_grid[-1]) + + def _ppf(q: Any, **_options: Any) -> NDArray[np.float64]: + q_arr = np.asarray(q, dtype=float) + q_clipped = np.clip(q_arr, cdf_lo, cdf_hi) + result: NDArray[np.float64] = interpolator(q_clipped) + # PchipInterpolator with extrapolate=False returns NaN outside the + # support; the explicit clip above prevents this, but keep a safety + # net for callers that pass q exactly on the grid endpoints. + return np.where(np.isnan(result), np.where(q_arr <= cdf_lo, x_lo, x_hi), result) + + return FittedComputationMethod[Any, Any]( + target=CharacteristicName.PPF, + sources=(CharacteristicName.CDF,), + func=cast(ComputationFunc[Any, Any], _ppf), + ) + + +__all__ = ["EmpiricalComputationStrategy"] diff --git a/src/pysatl_core/sampling/unuran/core/unuran_sampling_strategy.py b/src/pysatl_core/sampling/unuran/core/unuran_sampling_strategy.py index 33eb471..95ca54f 100644 --- a/src/pysatl_core/sampling/unuran/core/unuran_sampling_strategy.py +++ b/src/pysatl_core/sampling/unuran/core/unuran_sampling_strategy.py @@ -94,6 +94,30 @@ def sample(self, n: int, distr: Distribution, **options: Any) -> NumericArray: return self._sampler.sample(n) + def __deepcopy__(self, memo: dict[int, Any]) -> DefaultUnuranSamplingStrategy: + new = DefaultUnuranSamplingStrategy(config=self._config_value) + memo[id(self)] = new + return new + + def invalidate(self) -> None: + """ + Drop the cached UNURAN sampler. + + The next call to :meth:`sample` will rebuild a fresh sampler from the + distribution's current characteristics. Use this when the underlying + distribution state has changed (e.g. an empirical method has been + swapped) and the existing UNURAN generator was built on stale data — + keeping it would silently produce samples from the old distribution. + + Notes + ----- + Releases the only strong reference held by the strategy. Any external + code that captured ``self._sampler`` directly will keep its old sampler + alive and continue sampling from the previous distribution; such + references must be re-acquired by the caller. + """ + self._sampler = None + @property def config(self) -> UnuranMethodConfig: """Method configuration.""" diff --git a/tests/unit/distributions/empirical/__init__.py b/tests/unit/distributions/empirical/__init__.py new file mode 100644 index 0000000..ddc6388 --- /dev/null +++ b/tests/unit/distributions/empirical/__init__.py @@ -0,0 +1,11 @@ +""" +Empirical Distribution Tests +============================= + +Unit tests for EmpiricalDistribution and EmpiricalComputationStrategy, +covering indirection, method swapping, cache invalidation, and vectorised PPF. +""" + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" diff --git a/tests/unit/distributions/empirical/test_distribution.py b/tests/unit/distributions/empirical/test_distribution.py new file mode 100644 index 0000000..3b32807 --- /dev/null +++ b/tests/unit/distributions/empirical/test_distribution.py @@ -0,0 +1,265 @@ +""" +Unit tests for EmpiricalDistribution: with_method / set_method / indirection. +""" + +from __future__ import annotations + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" + +from typing import Any, cast + +import numpy as np +import pytest +from numpy.typing import NDArray + +from pysatl_core.distributions.empirical import ( + EmpiricalComputationStrategy, + EmpiricalDistribution, + FittedEmpirical, + ScipyGaussianKde, +) +from pysatl_core.types import CharacteristicName + + +class _ConstantFittedEmpirical: + def __init__(self, pdf_value: float, cdf_value: float) -> None: + self._pdf_value = pdf_value + self._cdf_value = cdf_value + self.fit_calls = 0 + + def pdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + x = np.atleast_1d(np.asarray(x, dtype=float)) + return np.full_like(x, self._pdf_value) + + def cdf(self, x: NDArray[np.float64]) -> NDArray[np.float64]: + x = np.atleast_1d(np.asarray(x, dtype=float)) + return np.full_like(x, self._cdf_value) + + +class _ConstantMethod: + def __init__(self, pdf_value: float, cdf_value: float) -> None: + self._pdf_value = pdf_value + self._cdf_value = cdf_value + self.fit_calls = 0 + + def fit(self, sample: NDArray[np.float64]) -> FittedEmpirical: + self.fit_calls += 1 + return _ConstantFittedEmpirical(self._pdf_value, self._cdf_value) + + +@pytest.fixture +def sample() -> NDArray[np.float64]: + rng = np.random.default_rng(42) + return rng.normal(0.0, 1.0, 300) + + +@pytest.fixture +def distr(sample: NDArray[np.float64]) -> EmpiricalDistribution: + return EmpiricalDistribution(sample) + + +class TestIndirection: + def test_pdf_indirection_reads_current_estimator(self, distr: EmpiricalDistribution) -> None: + distr._estimator = _ConstantFittedEmpirical(pdf_value=42.0, cdf_value=0.5) + result = distr.calculate_characteristic(CharacteristicName.PDF, np.array([0.0, 1.0])) + assert np.allclose(result, 42.0) + + def test_cdf_indirection_reads_current_estimator(self, distr: EmpiricalDistribution) -> None: + distr._estimator = _ConstantFittedEmpirical(pdf_value=1.0, cdf_value=0.7) + result = distr.calculate_characteristic(CharacteristicName.CDF, np.array([0.0, 2.0])) + assert np.allclose(result, 0.7) + + +class TestDefaultStrategy: + def test_default_computation_strategy_is_empirical(self, distr: EmpiricalDistribution) -> None: + assert isinstance(distr.computation_strategy, EmpiricalComputationStrategy) + + +class TestWithMethod: + def test_returns_new_instance(self, distr: EmpiricalDistribution) -> None: + clone = distr.with_method(ScipyGaussianKde(bandwidth="silverman")) + assert clone is not distr + assert isinstance(clone, EmpiricalDistribution) + + def test_shares_sample(self, distr: EmpiricalDistribution) -> None: + clone = distr.with_method(ScipyGaussianKde(bandwidth="silverman")) + assert clone.data is distr.data + + def test_changes_pdf(self, sample: NDArray[np.float64]) -> None: + scott = EmpiricalDistribution(sample, method=ScipyGaussianKde(bandwidth="scott")) + silverman = scott.with_method(ScipyGaussianKde(bandwidth="silverman")) + + x = np.array([0.0, 0.5, 1.0]) + pdf_scott = scott.calculate_characteristic(CharacteristicName.PDF, x) + pdf_silverman = silverman.calculate_characteristic(CharacteristicName.PDF, x) + + assert not np.allclose(pdf_scott, pdf_silverman) + + def test_preserves_original(self, distr: EmpiricalDistribution) -> None: + x = np.array([0.0, 1.0]) + pdf_before = distr.calculate_characteristic(CharacteristicName.PDF, x) + + distr.with_method(ScipyGaussianKde(bandwidth="silverman")) + + pdf_after = distr.calculate_characteristic(CharacteristicName.PDF, x) + assert np.allclose(pdf_before, pdf_after) + + def test_independent_strategies(self, distr: EmpiricalDistribution) -> None: + clone = distr.with_method(ScipyGaussianKde(bandwidth="silverman")) + assert clone.sampling_strategy is not distr.sampling_strategy + assert clone.computation_strategy is not distr.computation_strategy + + def test_method_property_reflects_new_method(self, distr: EmpiricalDistribution) -> None: + new_method = ScipyGaussianKde(bandwidth="silverman") + clone = distr.with_method(new_method) + assert clone.method is new_method + assert distr.method is not new_method + + def test_calls_fit_exactly_once_on_new_method(self, sample: NDArray[np.float64]) -> None: + method = _ConstantMethod(pdf_value=1.0, cdf_value=0.5) + distr = EmpiricalDistribution(sample) + + clone = distr.with_method(method) + + assert method.fit_calls == 1 + assert np.allclose( + clone.calculate_characteristic(CharacteristicName.PDF, np.array([0.0])), + 1.0, + ) + + def test_composes_with_sampling_strategy_override(self, distr: EmpiricalDistribution) -> None: + from pysatl_core.sampling.unuran.core.unuran_sampling_strategy import ( + DefaultUnuranSamplingStrategy, + ) + + new_sampling = DefaultUnuranSamplingStrategy() + chained = distr.with_method(ScipyGaussianKde(bandwidth="silverman")).with_sampling_strategy( + new_sampling + ) + assert chained.sampling_strategy is new_sampling + assert chained.method is not distr.method + + +class TestSetMethod: + def test_mutates_in_place_returns_none(self, distr: EmpiricalDistribution) -> None: + original_id = id(distr) + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + assert id(distr) == original_id + + def test_method_property_reflects_new_method(self, distr: EmpiricalDistribution) -> None: + new_method = ScipyGaussianKde(bandwidth="silverman") + distr.set_method(new_method) + assert distr.method is new_method + + def test_pdf_changes_after_swap(self, distr: EmpiricalDistribution) -> None: + x = np.array([0.0, 0.5, 1.0]) + pdf_before = distr.calculate_characteristic(CharacteristicName.PDF, x) + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + pdf_after = distr.calculate_characteristic(CharacteristicName.PDF, x) + assert not np.allclose(pdf_before, pdf_after) + + def test_estimator_object_replaced(self, distr: EmpiricalDistribution) -> None: + old_estimator = distr._estimator + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + assert distr._estimator is not old_estimator + + def test_calls_fit_exactly_once(self, distr: EmpiricalDistribution) -> None: + method = _ConstantMethod(pdf_value=7.0, cdf_value=0.3) + distr.set_method(method) + assert method.fit_calls == 1 + assert np.allclose( + distr.calculate_characteristic(CharacteristicName.PDF, np.array([0.0])), + 7.0, + ) + + def test_invalidates_sampling_sampler( + self, distr: EmpiricalDistribution, monkeypatch: pytest.MonkeyPatch + ) -> None: + class _StubSampler: + def __init__(self, distr: Any, config: Any) -> None: ... + + def sample(self, n: int) -> NDArray[np.float64]: + return np.zeros(n) + + monkeypatch.setattr( + "pysatl_core.sampling.unuran.core.unuran_sampling_strategy.DefaultUnuranSampler", + _StubSampler, + ) + _ = distr.sample(10) + from pysatl_core.sampling.unuran.core.unuran_sampling_strategy import ( + DefaultUnuranSamplingStrategy, + ) + + sampler_strategy = cast(DefaultUnuranSamplingStrategy, distr.sampling_strategy) + assert sampler_strategy._sampler is not None + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + assert sampler_strategy._sampler is None + + def test_computation_cache_cleared_via_estimator_tracking( + self, sample: NDArray[np.float64] + ) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + d = EmpiricalDistribution(sample, computation_strategy=strategy) + + from pysatl_core.distributions.computations.computation import FittedComputationMethod + + strategy.query_method(CharacteristicName.PDF, d) + strategy._ppf_cache[id(d)] = FittedComputationMethod[Any, Any]( + target=CharacteristicName.PPF, + sources=(CharacteristicName.CDF,), + func=lambda *a, **kw: 0.0, + ) + assert id(d) in strategy._ppf_cache + + d.set_method(ScipyGaussianKde(bandwidth="silverman")) + + strategy.query_method(CharacteristicName.PDF, d) + assert id(d) not in strategy._ppf_cache + + def test_sample_after_swap_rebuilds_sampler( + self, distr: EmpiricalDistribution, monkeypatch: pytest.MonkeyPatch + ) -> None: + built: list[Any] = [] + + class _StubSampler: + def __init__(self, distr: Any, config: Any) -> None: + built.append(self) + + def sample(self, n: int) -> NDArray[np.float64]: + return np.zeros(n) + + monkeypatch.setattr( + "pysatl_core.sampling.unuran.core.unuran_sampling_strategy.DefaultUnuranSampler", + _StubSampler, + ) + _ = distr.sample(5) + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + out = distr.sample(50) + + assert out.shape[0] == 50 + assert len(built) == 2 + assert built[0] is not built[1] + + def test_works_when_strategies_lack_invalidate(self, sample: NDArray[np.float64]) -> None: + class _BareSampling: + def sample(self, n: int, distr: Any, **options: Any) -> NDArray[np.float64]: + return np.zeros(n) + + class _BareComputation: + def query_method(self, state: Any, distr: Any, **options: Any) -> Any: + return distr.analytical_computations[CharacteristicName.PDF][ + next(iter(distr.analytical_computations[CharacteristicName.PDF])) + ] + + def explain_computation_path(self, state: Any, distr: Any) -> Any: + raise NotImplementedError + + d = EmpiricalDistribution( + sample, + sampling_strategy=_BareSampling(), + computation_strategy=_BareComputation(), # type: ignore[arg-type] + ) + + d.set_method(ScipyGaussianKde(bandwidth="silverman")) diff --git a/tests/unit/distributions/empirical/test_strategy.py b/tests/unit/distributions/empirical/test_strategy.py new file mode 100644 index 0000000..62b0cf2 --- /dev/null +++ b/tests/unit/distributions/empirical/test_strategy.py @@ -0,0 +1,212 @@ +""" +Unit tests for EmpiricalComputationStrategy. + +Covers: + - estimator-identity-aware cache invalidation + - degenerate behaviour for distributions without an _estimator attribute + - inheritance contract with DefaultComputationStrategy +""" + +from __future__ import annotations + +__author__ = "Artem Romanyuk" +__copyright__ = "Copyright (c) 2025 PySATL project" +__license__ = "SPDX-License-Identifier: MIT" + +from typing import Any, cast + +import numpy as np +import pytest + +from pysatl_core.distributions.computations.computation import FittedComputationMethod +from pysatl_core.distributions.empirical import EmpiricalComputationStrategy +from pysatl_core.distributions.strategies import DefaultComputationStrategy +from pysatl_core.types import CharacteristicName + + +class _FakeDistr: + def __init__(self, estimator: object | None) -> None: + self._estimator = estimator + + +def _put_dummy_ppf_cache_entry(strategy: EmpiricalComputationStrategy, fake_id: int = 0) -> None: + dummy = FittedComputationMethod[Any, Any]( + target=cast(CharacteristicName, "ppf"), + sources=(), + func=lambda *a, **kw: None, + ) + strategy._ppf_cache[fake_id] = dummy + + +class TestInheritance: + def test_is_default_strategy_subclass(self) -> None: + assert issubclass(EmpiricalComputationStrategy, DefaultComputationStrategy) + assert isinstance(EmpiricalComputationStrategy(), DefaultComputationStrategy) + + def test_caching_disabled_by_default(self) -> None: + strategy = EmpiricalComputationStrategy() + assert strategy.is_caching_enabled is False + + def test_caching_can_be_enabled(self) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + assert strategy.is_caching_enabled is True + + +class TestMaybeInvalidate: + def test_first_query_records_estimator_without_clearing(self) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + estimator = object() + distr = cast(Any, _FakeDistr(estimator)) + + strategy._maybe_invalidate(distr) + + assert strategy._tracked_estimator is estimator + assert strategy._cache == {} + assert strategy._ppf_cache == {} + + def test_same_estimator_keeps_cache(self) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + estimator = object() + distr = cast(Any, _FakeDistr(estimator)) + + strategy._maybe_invalidate(distr) + _put_dummy_ppf_cache_entry(strategy, fake_id=1) + strategy._maybe_invalidate(distr) + + assert 1 in strategy._ppf_cache + assert strategy._tracked_estimator is estimator + + def test_swapped_estimator_clears_cache(self) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + first = object() + second = object() + + strategy._maybe_invalidate(cast(Any, _FakeDistr(first))) + _put_dummy_ppf_cache_entry(strategy, fake_id=1) + _put_dummy_ppf_cache_entry(strategy, fake_id=2) + + strategy._maybe_invalidate(cast(Any, _FakeDistr(second))) + + assert strategy._ppf_cache == {} + assert strategy._tracked_estimator is second + + def test_distribution_without_estimator_attribute_does_not_crash(self) -> None: + strategy = EmpiricalComputationStrategy(enable_caching=True) + + class _Bare: + pass + + strategy._maybe_invalidate(cast(Any, _Bare())) + assert strategy._tracked_estimator is None + + _put_dummy_ppf_cache_entry(strategy, fake_id=0) + strategy._maybe_invalidate(cast(Any, _Bare())) + assert 0 in strategy._ppf_cache + + +class TestVectorizedPpf: + def _make_distr(self, sample_size: int = 300) -> Any: + np = pytest.importorskip("numpy") + from pysatl_core.distributions.empirical import EmpiricalDistribution + + rng = np.random.default_rng(0) + sample = rng.normal(0.0, 1.0, sample_size) + return EmpiricalDistribution(sample) + + def test_array_input_returns_array(self) -> None: + np = pytest.importorskip("numpy") + distr = self._make_distr() + q = np.linspace(0.05, 0.95, 50) + result = distr.calculate_characteristic(CharacteristicName.PPF, q) + assert result.shape == (50,) + + def test_result_is_monotonic_nondecreasing(self) -> None: + np = pytest.importorskip("numpy") + distr = self._make_distr() + q = np.linspace(0.01, 0.99, 200) + result = distr.calculate_characteristic(CharacteristicName.PPF, q) + assert np.all(np.diff(result) >= -1e-10) + + def test_scalar_input_returns_scalar(self) -> None: + distr = self._make_distr() + result = distr.calculate_characteristic(CharacteristicName.PPF, 0.5) + assert np.ndim(result) == 0 + + def test_matches_root_finding_within_tolerance(self) -> None: + np = pytest.importorskip("numpy") + from pysatl_core.distributions.empirical import EmpiricalDistribution + from pysatl_core.distributions.strategies import DefaultComputationStrategy + + rng = np.random.default_rng(7) + sample = rng.normal(0.0, 1.0, 400) + + fast = EmpiricalDistribution(sample) + slow = EmpiricalDistribution(sample, computation_strategy=DefaultComputationStrategy()) + quantiles = np.linspace(0.1, 0.9, 9) + fast_vals = fast.calculate_characteristic(CharacteristicName.PPF, quantiles) + slow_vals = np.array( + [ + float(np.squeeze(slow.calculate_characteristic(CharacteristicName.PPF, float(q)))) + for q in quantiles + ] + ) + assert np.allclose(fast_vals, slow_vals, atol=5e-2) + + def test_handles_extreme_quantiles_without_nan(self) -> None: + np = pytest.importorskip("numpy") + distr = self._make_distr() + q = np.array([1e-7, 0.5, 1.0 - 1e-7]) + result = distr.calculate_characteristic(CharacteristicName.PPF, q) + assert not np.any(np.isnan(result)) + + def test_ppf_recomputed_after_estimator_swap(self) -> None: + np = pytest.importorskip("numpy") + from pysatl_core.distributions.empirical import ( + EmpiricalDistribution, + ScipyGaussianKde, + ) + + rng = np.random.default_rng(3) + sample = rng.normal(0.0, 1.0, 300) + strategy = EmpiricalComputationStrategy(enable_caching=True) + distr = EmpiricalDistribution(sample, computation_strategy=strategy) + + first = distr.calculate_characteristic(CharacteristicName.PPF, 0.5) + cached_after_first = strategy._ppf_cache.get(id(distr)) + assert cached_after_first is not None + + distr.set_method(ScipyGaussianKde(bandwidth="silverman")) + second = distr.calculate_characteristic(CharacteristicName.PPF, 0.5) + cached_after_swap = strategy._ppf_cache.get(id(distr)) + + assert cached_after_swap is not None + assert cached_after_swap is not cached_after_first + assert np.isfinite(first) + assert np.isfinite(second) + + +class TestQueryMethodIntegration: + def test_estimator_swap_clears_fitted_cache_via_query_method(self) -> None: + np = pytest.importorskip("numpy") + + from pysatl_core.distributions.empirical import ( + EmpiricalDistribution, + ScipyGaussianKde, + ) + + rng = np.random.default_rng(0) + sample = rng.normal(0, 1, 200) + strategy = EmpiricalComputationStrategy(enable_caching=True) + distr = EmpiricalDistribution(sample, computation_strategy=strategy) + + strategy.query_method(CharacteristicName.PDF, distr) + _put_dummy_ppf_cache_entry(strategy, fake_id=99999) + assert 99999 in strategy._ppf_cache + + new_estimator = ScipyGaussianKde(bandwidth="silverman").fit(sample) + distr._estimator = new_estimator + + strategy.query_method(CharacteristicName.PDF, distr) + + assert 99999 not in strategy._ppf_cache + assert strategy._tracked_estimator is new_estimator diff --git a/tests/unit/sampling/unuran/core/test_unuran_sampling_strategy.py b/tests/unit/sampling/unuran/core/test_unuran_sampling_strategy.py index 92fb6a8..2eb8600 100644 --- a/tests/unit/sampling/unuran/core/test_unuran_sampling_strategy.py +++ b/tests/unit/sampling/unuran/core/test_unuran_sampling_strategy.py @@ -177,6 +177,45 @@ def test_sampler_is_reused_across_multiple_calls(self, monkeypatch: pytest.Monke assert sampler_first is sampler_second + def test_invalidate_clears_cached_sampler(self, monkeypatch: pytest.MonkeyPatch) -> None: + """invalidate() drops the cached sampler so the next call rebuilds it.""" + monkeypatch.setattr(_SAMPLER_MODULE, _StubSampler) + strategy = DefaultUnuranSamplingStrategy() + distr = _make_continuous_distr() + + strategy.sample(2, distr) + assert strategy._sampler is not None + + strategy.invalidate() + assert strategy._sampler is None + + def test_invalidate_forces_new_sampler_on_next_sample( + self, monkeypatch: pytest.MonkeyPatch + ) -> None: + """After invalidate(), the next sample() builds a fresh sampler instance.""" + monkeypatch.setattr(_SAMPLER_MODULE, _StubSampler) + strategy = DefaultUnuranSamplingStrategy() + distr = _make_continuous_distr() + + strategy.sample(2, distr) + first_sampler = strategy._sampler + + strategy.invalidate() + strategy.sample(2, distr) + second_sampler = strategy._sampler + + assert first_sampler is not None + assert second_sampler is not None + assert first_sampler is not second_sampler + + def test_invalidate_is_idempotent_when_no_sampler_cached(self) -> None: + """invalidate() on a fresh strategy with no sampler is a no-op and does not raise.""" + strategy = DefaultUnuranSamplingStrategy() + assert strategy._sampler is None + + strategy.invalidate() + assert strategy._sampler is None + def test_falls_back_to_default_strategy_when_sampler_init_fails( self, monkeypatch: pytest.MonkeyPatch ) -> None: