ecliseutils 0.1.0__tar.gz

ecliseutils-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Wentinn Liao
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ecliseutils-0.1.0/PKG-INFO

@@ -0,0 +1,105 @@
+Metadata-Version: 2.4
+Name: ecliseutils
+Version: 0.1.0
+Summary: Shared general-purpose research utilities (tensor/module batching, linear algebra, ARE/ODE solvers, labeled arrays, plotting) used across multiple projects.
+Author: Wentinn Liao
+License: MIT
+Project-URL: Homepage, https://github.com/JophiArcana/ecliseutils
+Project-URL: Repository, https://github.com/JophiArcana/ecliseutils
+Project-URL: Issues, https://github.com/JophiArcana/ecliseutils/issues
+Keywords: pytorch,numpy,linear-algebra,research-utilities,vmap,tensordict
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: torch
+Requires-Dist: tensordict
+Requires-Dist: einops
+Requires-Dist: scipy
+Requires-Dist: matplotlib
+Requires-Dist: tqdm
+Requires-Dist: pandas
+Provides-Extra: ode
+Requires-Dist: torchdiffeq; extra == "ode"
+Dynamic: license-file
+
+# ecliseutils
+
+Shared general-purpose research utilities used across multiple projects (e.g.
+`KF_RNN`, `event_camera`). This package is the single source of truth for code
+that previously lived as diverged copies inside each project's
+`infrastructure/utils.py`.
+
+## What's here
+
+| Module | Contents |
+|--------|----------|
+| `settings` | `configure(device, dtype, precision, seed, debug)`, `default_dtype`, safe-global registration. No hardcoded paths; no import-time torch mutation. |
+| `types` | `ModelPair = tuple[nn.Module, TensorDict]` and shared aliases. |
+| `labeled_array` | `LabeledArray` / `LabeledDataset` (in-house replacement for `dimarray`). |
+| `modules` | Module/tensor/TensorDict batching: `stack_module_arr`, `run_module_arr`, `multi_vmap`, `td_items`, ... |
+| `arrays` | NumPy object-array comprehensions: `multi_map`, `multi_zip`, `broadcast_dim_arrays`, ... |
+| `linalg` | `pow_series`, `batch_trace`, `kl_div`, `sqrtm`, `hadamard_conjugation*`, `inverse`, `eig_some`, ... |
+| `are` | `solve_discrete_are` (torch Schur) and `solve_continuous_are` (scipy `ordqz` fallback) + residual tests. |
+| `ode` | `batch_odeint`, `linspace`, `geomspace` (lazy `torchdiffeq`). |
+| `ensemble` | `EnsembleModule` chunked vmap wrapper. |
+| `fast_conv_scan` | `conv_scan` exponential cumulative scan autograd op. |
+| `recursive` | Dotted-path `rgetattr` / `rsetattr` / `rgetitem` / `rsetitem`. |
+| `dicts` | `flatten_nested_dict`, `map_dict`, `nested_type`, `call_func_with_kwargs`. |
+| `io` | `torch_load`, `empty_cache`, `reset_seed`, `model_size`. |
+| `memory` | `get_tensors_in_memory`, `track_tensor_diff`, ... |
+| `timing` | `Timer`, `print_disabled`, `print_enabled`, `track_calls`. |
+| `plotting` | `color`, `confidence_ellipse`. |
+
+## Install
+
+From PyPI (once published):
+
+```bash
+pip install ecliseutils
+# with ODE support:
+pip install "ecliseutils[ode]"
+```
+
+From GitHub (works on any device, no clone needed):
+
+```bash
+pip install "git+https://github.com/JophiArcana/ecliseutils.git"
+# pin to a release:
+pip install "git+https://github.com/JophiArcana/ecliseutils.git@v0.1.0"
+# with ODE support:
+pip install "ecliseutils[ode] @ git+https://github.com/JophiArcana/ecliseutils.git"
+```
+
+Editable (for local development):
+
+```bash
+pip install -e /path/to/ecliseutils
+# with ODE support:
+pip install -e "/path/to/ecliseutils[ode]"
+```
+
+## Usage
+
+```python
+import torch
+import ecliseutils as eu
+
+# Configure once at startup (typically from your project's settings module).
+eu.configure(device="cpu", dtype=torch.float64, precision=8, seed=1212)
+
+P = eu.solve_discrete_are(A, B, Q, R)
+```
+
+`torchdiffeq` is only required if you call `ecliseutils.ode.batch_odeint`; it is
+imported lazily.

ecliseutils-0.1.0/README.md

@@ -0,0 +1,70 @@
+# ecliseutils
+
+Shared general-purpose research utilities used across multiple projects (e.g.
+`KF_RNN`, `event_camera`). This package is the single source of truth for code
+that previously lived as diverged copies inside each project's
+`infrastructure/utils.py`.
+
+## What's here
+
+| Module | Contents |
+|--------|----------|
+| `settings` | `configure(device, dtype, precision, seed, debug)`, `default_dtype`, safe-global registration. No hardcoded paths; no import-time torch mutation. |
+| `types` | `ModelPair = tuple[nn.Module, TensorDict]` and shared aliases. |
+| `labeled_array` | `LabeledArray` / `LabeledDataset` (in-house replacement for `dimarray`). |
+| `modules` | Module/tensor/TensorDict batching: `stack_module_arr`, `run_module_arr`, `multi_vmap`, `td_items`, ... |
+| `arrays` | NumPy object-array comprehensions: `multi_map`, `multi_zip`, `broadcast_dim_arrays`, ... |
+| `linalg` | `pow_series`, `batch_trace`, `kl_div`, `sqrtm`, `hadamard_conjugation*`, `inverse`, `eig_some`, ... |
+| `are` | `solve_discrete_are` (torch Schur) and `solve_continuous_are` (scipy `ordqz` fallback) + residual tests. |
+| `ode` | `batch_odeint`, `linspace`, `geomspace` (lazy `torchdiffeq`). |
+| `ensemble` | `EnsembleModule` chunked vmap wrapper. |
+| `fast_conv_scan` | `conv_scan` exponential cumulative scan autograd op. |
+| `recursive` | Dotted-path `rgetattr` / `rsetattr` / `rgetitem` / `rsetitem`. |
+| `dicts` | `flatten_nested_dict`, `map_dict`, `nested_type`, `call_func_with_kwargs`. |
+| `io` | `torch_load`, `empty_cache`, `reset_seed`, `model_size`. |
+| `memory` | `get_tensors_in_memory`, `track_tensor_diff`, ... |
+| `timing` | `Timer`, `print_disabled`, `print_enabled`, `track_calls`. |
+| `plotting` | `color`, `confidence_ellipse`. |
+
+## Install
+
+From PyPI (once published):
+
+```bash
+pip install ecliseutils
+# with ODE support:
+pip install "ecliseutils[ode]"
+```
+
+From GitHub (works on any device, no clone needed):
+
+```bash
+pip install "git+https://github.com/JophiArcana/ecliseutils.git"
+# pin to a release:
+pip install "git+https://github.com/JophiArcana/ecliseutils.git@v0.1.0"
+# with ODE support:
+pip install "ecliseutils[ode] @ git+https://github.com/JophiArcana/ecliseutils.git"
+```
+
+Editable (for local development):
+
+```bash
+pip install -e /path/to/ecliseutils
+# with ODE support:
+pip install -e "/path/to/ecliseutils[ode]"
+```
+
+## Usage
+
+```python
+import torch
+import ecliseutils as eu
+
+# Configure once at startup (typically from your project's settings module).
+eu.configure(device="cpu", dtype=torch.float64, precision=8, seed=1212)
+
+P = eu.solve_discrete_are(A, B, Q, R)
+```
+
+`torchdiffeq` is only required if you call `ecliseutils.ode.batch_odeint`; it is
+imported lazily.

ecliseutils-0.1.0/pyproject.toml

@@ -0,0 +1,54 @@
+[build-system]
+requires = ["setuptools>=68", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "ecliseutils"
+version = "0.1.0"
+description = "Shared general-purpose research utilities (tensor/module batching, linear algebra, ARE/ODE solvers, labeled arrays, plotting) used across multiple projects."
+readme = "README.md"
+license = { text = "MIT" }
+requires-python = ">=3.9"
+authors = [{ name = "Wentinn Liao" }]
+keywords = [
+    "pytorch",
+    "numpy",
+    "linear-algebra",
+    "research-utilities",
+    "vmap",
+    "tensordict",
+]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.9",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Scientific/Engineering",
+]
+dependencies = [
+    "numpy",
+    "torch",
+    "tensordict",
+    "einops",
+    "scipy",
+    "matplotlib",
+    "tqdm",
+    "pandas",
+]
+
+[project.optional-dependencies]
+# Required only for the ODE integration helpers in ``ecliseutils.ode``.
+ode = ["torchdiffeq"]
+
+[project.urls]
+Homepage = "https://github.com/JophiArcana/ecliseutils"
+Repository = "https://github.com/JophiArcana/ecliseutils"
+Issues = "https://github.com/JophiArcana/ecliseutils/issues"
+
+[tool.setuptools.packages.find]
+where = ["src"]

ecliseutils-0.1.0/setup.cfg

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

ecliseutils-0.1.0/src/ecliseutils/__init__.py

@@ -0,0 +1,99 @@
+"""ecliseutils: shared general-purpose research utilities.
+
+Submodules group the utilities by concern (``modules``, ``arrays``, ``linalg``,
+``are``, ``ode``, ``labeled_array``, ``ensemble``, ``fast_conv_scan``,
+``recursive``, ``dicts``, ``io``, ``memory``, ``timing``, ``plotting``,
+``settings``, ``types``). The most commonly used names are re-exported here for
+convenience (``import ecliseutils as eu; eu.stack_module_arr(...)``).
+
+Projects should call :func:`ecliseutils.configure` early (typically from their
+own ``settings`` module) to set device/dtype/precision/seed.
+"""
+
+from __future__ import annotations
+
+from . import (
+    arrays,
+    are,
+    dicts,
+    ensemble,
+    fast_conv_scan,
+    io,
+    labeled_array,
+    linalg,
+    memory,
+    modules,
+    ode,
+    plotting,
+    recursive,
+    settings,
+    timing,
+    types,
+)
+
+from .settings import configure, set_debug, register_safe_globals, default_dtype
+from .types import ModelPair
+
+from .labeled_array import LabeledArray, LabeledDataset, array_of, put_object, as_ndarray
+from .modules import (
+    stack_tensor_arr,
+    stack_module_arr,
+    stack_module_arr_preserve_reference,
+    run_module_arr,
+    multi_vmap,
+    buffer_dict,
+    td_items,
+    td_get,
+    parameter_td,
+    mask_dataset_with_total_sequence_length,
+    broadcast_shapes,
+    get_all_hooks,
+)
+from .arrays import (
+    multi_iter,
+    multi_enumerate,
+    multi_map,
+    multi_zip,
+    dim_array_like,
+    broadcast_dim_array_shapes,
+    broadcast_dim_arrays,
+    take_from_dim_array,
+)
+from .linalg import (
+    pow_series,
+    batch_trace,
+    kl_div,
+    sqrtm,
+    complex,
+    ceildiv,
+    ceil,
+    T,
+    hadamard_conjugation,
+    hadamard_conjugation_diff_order1,
+    hadamard_conjugation_diff_order2,
+    inverse,
+    eig_some,
+)
+from .are import (
+    solve_discrete_are,
+    solve_continuous_are,
+    test_discrete_are,
+    test_continuous_are,
+)
+from .ode import linspace, geomspace, batch_odeint
+from .ensemble import EnsembleModule, DEFAULT_SPLIT_SIZE
+from .fast_conv_scan import ConvScanFn, conv_scan
+from .recursive import rgetattr, rsetattr, rhasattr, rgetitem, rsetitem
+from .dicts import flatten_nested_dict, map_dict, nested_type, print_dict, call_func_with_kwargs, hash_hex
+from .io import torch_load, empty_cache, reset_seed, model_size
+from .memory import (
+    get_tensors_in_memory,
+    print_tensors_in_memory,
+    get_tensors_in_memory_shape,
+    track_tensor_diff,
+)
+from .timing import Timer, identity, PTR, print_disabled, print_enabled, track_calls
+from .plotting import color, confidence_ellipse
+
+
+__version__ = "0.1.0"