sphncs 0.3.1__tar.gz

sphncs-0.3.1/LICENSE

@@ -0,0 +1,178 @@
+Copyright 2026 Shawn Davis
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS

sphncs-0.3.1/PKG-INFO

@@ -0,0 +1,169 @@
+Metadata-Version: 2.4
+Name: sphncs
+Version: 0.3.1
+Summary: Similarity-Preserving Hierarchical Nonparametric Clustering System
+Author: Shawn Davis
+License-Expression: Apache-2.0
+Project-URL: Homepage, https://github.com/shawn-davis/sphncs
+Project-URL: Repository, https://github.com/shawn-davis/sphncs
+Project-URL: Issues, https://github.com/shawn-davis/sphncs/issues
+Keywords: clustering,density-estimation,fastmap,log-analysis,machine-learning
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy>=1.24
+Requires-Dist: scipy>=1.10
+Requires-Dist: scikit-learn>=1.3
+Requires-Dist: KDEpy>=1.1
+Requires-Dist: FastMapy>=0.2.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7.4; extra == "dev"
+Requires-Dist: ruff>=0.4; extra == "dev"
+Provides-Extra: bench
+Requires-Dist: rapidfuzz>=3.14; extra == "bench"
+Requires-Dist: matplotlib>=3.8; extra == "bench"
+Requires-Dist: drain3>=0.9.11; extra == "bench"
+Dynamic: license-file
+
+# sphncs
+
+Licensed under [Apache-2.0](LICENSE).
+
+`sphncs` is the **Similarity-Preserving Hierarchical Nonparametric Clustering
+System**. It clusters any objects with a well-defined, non-negative distance
+metric by embedding on-demand distances with FastMap and finding density
+separations with `KDEpy.FFTKDE`.
+
+It supports a lightweight single-dimension mode and an optional spectral-consensus
+mode that combines KDE cluster assignments from several FastMap dimensions. An
+optional first KDE can partition objects with any user-supplied scalar feature.
+In spectral-consensus mode, automatic `k` is the median number of clusters
+observed by the per-dimension KDE fits, rather than the number of embeddings.
+Set `consensus_n_clusters` to `"min"`, `"mean"`, `"median"`, or `"max"` to
+select another reduction of those observed counts; an explicit integer remains
+available when a fixed target is required.
+`extrema_prominence_fraction` controls how deep a KDE valley must be relative
+to that KDE's density range (the default is `0.05`); smaller fractions preserve
+more candidate modes.
+The estimator uses three FastMap pivot-refinement passes. Configure the number
+of passes with `fastmap_iters`.
+
+```python
+from dataclasses import dataclass
+
+from sphncs import SphncsClusterer
+
+
+@dataclass
+class Point:
+    x: float
+
+
+def distance(left: Point, right: Point) -> float:
+    return abs(left.x - right.x)
+
+model = SphncsClusterer(metric=distance)
+labels = model.fit_predict([Point(0.0), Point(0.2), Point(10.0), Point(10.2)])
+print(model.representatives_)
+```
+
+For optional first-stage partitioning, provide a scalar feature. Partitioning is
+hard routing: objects in different feature intervals are clustered separately.
+
+```python
+model = SphncsClusterer(
+    metric=distance,
+    partitioning=True,
+    partitioning_feature=lambda point: point.x,
+)
+```
+
+String distances such as `normalized_levenshtein` and `char_ngram_jaccard`
+remain available in `sphncs.distances`. They are conveniences, not a restriction
+on the estimator's input type.
+
+### Log clustering
+
+`LogSPHNCS` is the log-specific adapter. Its `log_filters` replace selected
+fields with filter-specific, fixed-width masks before embedding and clustering.
+The optional initial partitioner uses filtered strings by default; pass
+`partitioning_before_transform=True` to derive its feature from the original
+raw strings first. It uses length by default;
+set `partitioning_feature="entropy"` for character Shannon entropy or
+`"normalized_entropy"` for character-use evenness. Every marker is four
+characters long: timestamps use `<#T>`, severity uses `<#S>`, UUIDs use `<#U>`,
+IPs use `<#I>`, hex values use `<#H>`, numbers use `<#N>`, paths use `<#P>`,
+quoted values use `<#Q>`, and identifiers use `<#D>`. Choose individual filters
+(`timestamp`, `severity`, `uuid`, `ip`, `hex`, `number`, `path`, `quoted`, and
+`identifier`), use `variable` for all value-masking filters, or use `all` for
+every filter.
+
+The `hex` filter recognizes `0x`-prefixed values. Short bare hexadecimal
+sequence fields such as `0000000e` are normalized by the `number` filter, so
+they match their decimal-only counterparts without masking longer component IDs.
+
+`representatives_` contains the filtered form by default. The original training
+lines remain available in the position-aligned `raw_strings_`, through
+`get_raw_string(index)`, and as `raw_representatives_` for the cluster
+representatives.
+
+```python
+model = LogSPHNCS(
+    log_filters=["timestamp", "severity", "variable"],
+)
+```
+
+### LogSPHNCS
+
+`LogSPHNCS` defaults to all log filters,
+4-gram Jaccard, length partitioning, 10-dimensional spectral consensus, and
+exact filtered-template compression. Raw records remain aligned in
+`raw_strings_` and `raw_representatives_`.
+
+```python
+from sphncs import LogSPHNCS
+
+model = LogSPHNCS().fit(log_lines)
+```
+
+### Model persistence
+
+Fitted models can be saved to a versioned, integrity-checked `.sphncs` archive
+and loaded later. The archive preserves prediction and transformation state.
+Each FastMap projection is persisted with FastMapy’s native versioned format.
+
+```python
+model.save("logs.sphncs")
+restored = LogSPHNCS.load("logs.sphncs")
+assert restored.predict(log_lines).tolist() == model.predict(log_lines).tolist()
+```
+
+Archives use Python pickle to support arbitrary input objects and user-supplied
+metrics or transformers. Only load archives from sources you trust. Callables
+must be importable functions (rather than lambdas or nested functions) to save
+reliably. Format version 2 is validated on load; unsupported future formats are
+rejected rather than loaded incorrectly.
+
+`fastmapy` is included as the `vendor/fastmapy` Git submodule and isolated behind
+an embedding adapter. It is used for all FastMap fits, including its `fit_many`
+API for spectral-consensus embeddings; no pairwise string-distance matrix is
+created or retained.
+
+## Development
+
+```bash
+git submodule update --init --recursive
+python -m pip install -e vendor/fastmapy
+python -m pip install -e '.[dev]'
+python -m pytest
+```

sphncs-0.3.1/README.md

@@ -0,0 +1,132 @@
+# sphncs
+
+Licensed under [Apache-2.0](LICENSE).
+
+`sphncs` is the **Similarity-Preserving Hierarchical Nonparametric Clustering
+System**. It clusters any objects with a well-defined, non-negative distance
+metric by embedding on-demand distances with FastMap and finding density
+separations with `KDEpy.FFTKDE`.
+
+It supports a lightweight single-dimension mode and an optional spectral-consensus
+mode that combines KDE cluster assignments from several FastMap dimensions. An
+optional first KDE can partition objects with any user-supplied scalar feature.
+In spectral-consensus mode, automatic `k` is the median number of clusters
+observed by the per-dimension KDE fits, rather than the number of embeddings.
+Set `consensus_n_clusters` to `"min"`, `"mean"`, `"median"`, or `"max"` to
+select another reduction of those observed counts; an explicit integer remains
+available when a fixed target is required.
+`extrema_prominence_fraction` controls how deep a KDE valley must be relative
+to that KDE's density range (the default is `0.05`); smaller fractions preserve
+more candidate modes.
+The estimator uses three FastMap pivot-refinement passes. Configure the number
+of passes with `fastmap_iters`.
+
+```python
+from dataclasses import dataclass
+
+from sphncs import SphncsClusterer
+
+
+@dataclass
+class Point:
+    x: float
+
+
+def distance(left: Point, right: Point) -> float:
+    return abs(left.x - right.x)
+
+model = SphncsClusterer(metric=distance)
+labels = model.fit_predict([Point(0.0), Point(0.2), Point(10.0), Point(10.2)])
+print(model.representatives_)
+```
+
+For optional first-stage partitioning, provide a scalar feature. Partitioning is
+hard routing: objects in different feature intervals are clustered separately.
+
+```python
+model = SphncsClusterer(
+    metric=distance,
+    partitioning=True,
+    partitioning_feature=lambda point: point.x,
+)
+```
+
+String distances such as `normalized_levenshtein` and `char_ngram_jaccard`
+remain available in `sphncs.distances`. They are conveniences, not a restriction
+on the estimator's input type.
+
+### Log clustering
+
+`LogSPHNCS` is the log-specific adapter. Its `log_filters` replace selected
+fields with filter-specific, fixed-width masks before embedding and clustering.
+The optional initial partitioner uses filtered strings by default; pass
+`partitioning_before_transform=True` to derive its feature from the original
+raw strings first. It uses length by default;
+set `partitioning_feature="entropy"` for character Shannon entropy or
+`"normalized_entropy"` for character-use evenness. Every marker is four
+characters long: timestamps use `<#T>`, severity uses `<#S>`, UUIDs use `<#U>`,
+IPs use `<#I>`, hex values use `<#H>`, numbers use `<#N>`, paths use `<#P>`,
+quoted values use `<#Q>`, and identifiers use `<#D>`. Choose individual filters
+(`timestamp`, `severity`, `uuid`, `ip`, `hex`, `number`, `path`, `quoted`, and
+`identifier`), use `variable` for all value-masking filters, or use `all` for
+every filter.
+
+The `hex` filter recognizes `0x`-prefixed values. Short bare hexadecimal
+sequence fields such as `0000000e` are normalized by the `number` filter, so
+they match their decimal-only counterparts without masking longer component IDs.
+
+`representatives_` contains the filtered form by default. The original training
+lines remain available in the position-aligned `raw_strings_`, through
+`get_raw_string(index)`, and as `raw_representatives_` for the cluster
+representatives.
+
+```python
+model = LogSPHNCS(
+    log_filters=["timestamp", "severity", "variable"],
+)
+```
+
+### LogSPHNCS
+
+`LogSPHNCS` defaults to all log filters,
+4-gram Jaccard, length partitioning, 10-dimensional spectral consensus, and
+exact filtered-template compression. Raw records remain aligned in
+`raw_strings_` and `raw_representatives_`.
+
+```python
+from sphncs import LogSPHNCS
+
+model = LogSPHNCS().fit(log_lines)
+```
+
+### Model persistence
+
+Fitted models can be saved to a versioned, integrity-checked `.sphncs` archive
+and loaded later. The archive preserves prediction and transformation state.
+Each FastMap projection is persisted with FastMapy’s native versioned format.
+
+```python
+model.save("logs.sphncs")
+restored = LogSPHNCS.load("logs.sphncs")
+assert restored.predict(log_lines).tolist() == model.predict(log_lines).tolist()
+```
+
+Archives use Python pickle to support arbitrary input objects and user-supplied
+metrics or transformers. Only load archives from sources you trust. Callables
+must be importable functions (rather than lambdas or nested functions) to save
+reliably. Format version 2 is validated on load; unsupported future formats are
+rejected rather than loaded incorrectly.
+
+`fastmapy` is included as the `vendor/fastmapy` Git submodule and isolated behind
+an embedding adapter. It is used for all FastMap fits, including its `fit_many`
+API for spectral-consensus embeddings; no pairwise string-distance matrix is
+created or retained.
+
+## Development
+
+```bash
+git submodule update --init --recursive
+python -m pip install -e vendor/fastmapy
+python -m pip install -e '.[dev]'
+python -m pytest
+```

sphncs-0.3.1/pyproject.toml

@@ -0,0 +1,47 @@
+[build-system]
+requires = ["setuptools>=68"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "sphncs"
+version = "0.3.1"
+description = "Similarity-Preserving Hierarchical Nonparametric Clustering System"
+readme = "README.md"
+requires-python = ">=3.10"
+license = "Apache-2.0"
+authors = [{ name = "Shawn Davis" }]
+keywords = ["clustering", "density-estimation", "fastmap", "log-analysis", "machine-learning"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Science/Research",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3 :: Only",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]
+dependencies = [
+    "numpy>=1.24",
+    "scipy>=1.10",
+    "scikit-learn>=1.3",
+    "KDEpy>=1.1",
+    "FastMapy>=0.2.0",
+]
+
+[project.optional-dependencies]
+dev = ["pytest>=7.4", "ruff>=0.4"]
+bench = ["rapidfuzz>=3.14", "matplotlib>=3.8", "drain3>=0.9.11"]
+
+[project.urls]
+Homepage = "https://github.com/shawn-davis/sphncs"
+Repository = "https://github.com/shawn-davis/sphncs"
+Issues = "https://github.com/shawn-davis/sphncs/issues"
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+[tool.setuptools.packages.find]
+include = ["sphncs*"]

sphncs-0.3.1/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

sphncs-0.3.1/sphncs/__init__.py

@@ -0,0 +1,8 @@
+"""Similarity-Preserving Hierarchical Nonparametric Clustering System."""
+
+from .estimator import SphncsClusterer
+from .logs import LogSPHNCS
+from .persistence import ModelPersistenceError
+from .preprocessing import LogPreprocessor
+
+__all__ = ["LogPreprocessor", "LogSPHNCS", "ModelPersistenceError", "SphncsClusterer"]

sphncs-0.3.1/sphncs/density.py

@@ -0,0 +1,81 @@
+"""One-dimensional KDE evaluation, extrema, labels, and representatives."""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+import numpy as np
+from scipy.signal import find_peaks
+
+
+@dataclass
+class DensityModel:
+    grid: np.ndarray
+    density: np.ndarray
+    boundaries: np.ndarray
+    modes: np.ndarray
+    labels: np.ndarray
+    representative_indices: np.ndarray
+
+    def predict(self, values: np.ndarray) -> np.ndarray:
+        return np.searchsorted(self.boundaries, values, side="right").astype(int)
+
+
+def fit_density(
+    values: np.ndarray,
+    *,
+    bandwidth: str | float = "ISJ",
+    grid_points: int = 1024,
+    prominence: float | None = None,
+    prominence_fraction: float = 0.05,
+    min_samples: int = 3,
+) -> DensityModel:
+    """Fit FFTKDE and turn meaningful minima into interval labels."""
+    values = np.asarray(values, dtype=float).reshape(-1)
+    if values.size == 0:
+        raise ValueError("KDE requires at least one value")
+    lower, upper = float(values.min()), float(values.max())
+    if values.size < min_samples or np.isclose(lower, upper):
+        grid = np.linspace(lower - 0.5, upper + 0.5 if upper >= lower else lower + 0.5, max(2, grid_points))
+        density = np.zeros_like(grid)
+        mode = np.array([float(values.mean())])
+        labels = np.zeros(values.size, dtype=int)
+        return DensityModel(grid, density, np.array([], dtype=float), mode, labels, np.array([int(np.argmin(abs(values - mode[0])))]) )
+
+    padding = max((upper - lower) * 0.1, np.finfo(float).eps * 100)
+    grid = np.linspace(lower - padding, upper + padding, grid_points)
+    try:
+        from KDEpy import FFTKDE
+    except ImportError as exc:  # pragma: no cover - dependency declaration is authoritative
+        raise ImportError("KDEpy is required; install sphncs with its runtime dependencies.") from exc
+    try:
+        density = np.asarray(FFTKDE(kernel="gaussian", bw=bandwidth).fit(values).evaluate(grid), dtype=float)
+    except ValueError:
+        # ISJ can fail to find a root for small, highly discrete partitions.
+        # Keep an explicit user-selected bandwidth strict, but make the default
+        # automatic choice robust by falling back to another KDEpy rule.
+        if not isinstance(bandwidth, str) or bandwidth.upper() != "ISJ":
+            raise
+        density = np.asarray(FFTKDE(kernel="gaussian", bw="silverman").fit(values).evaluate(grid), dtype=float)
+    scale = float(np.ptp(density))
+    effective_prominence = (
+        prominence
+        if prominence is not None
+        else max(scale * prominence_fraction, np.finfo(float).eps)
+    )
+    minima, _ = find_peaks(-density, prominence=effective_prominence)
+    boundaries = grid[minima]
+    labels = np.searchsorted(boundaries, values, side="right").astype(int)
+
+    modes: list[float] = []
+    representatives: list[int] = []
+    for label in range(len(boundaries) + 1):
+        left = -np.inf if label == 0 else boundaries[label - 1]
+        right = np.inf if label == len(boundaries) else boundaries[label]
+        mask = (grid > left) & (grid <= right)
+        mode = float(grid[mask][np.argmax(density[mask])])
+        members = np.flatnonzero(labels == label)
+        if members.size:
+            modes.append(mode)
+            representatives.append(int(members[np.argmin(abs(values[members] - mode))]))
+    return DensityModel(grid, density, boundaries, np.asarray(modes), labels, np.asarray(representatives, dtype=int))