commkit 1.0.0__py3-none-any.whl

commkit/equalization/__init__.py

@@ -0,0 +1,49 @@
+"""
+Adaptive and block channel equalization.
+
+This package provides sequential (LMS, RLS, CMA, RDE) and block / frequency-domain
+equalizers with optional carrier-phase recovery, plus linear (zero-forcing) and
+pilot-tone polarization-demultiplexing routines. Numba (CPU) and JAX (GPU) kernel
+backends dispatch automatically based on where the input data resides.
+
+The public API is unchanged from when this was a single module:
+``from commkit.equalization import lms, rls, cma, rde, ...`` continues to work.
+"""
+
+from __future__ import annotations
+
+# Re-exported for tests/benchmarks that reach package internals through this
+# namespace (historically ``equalization._get_jax`` etc.). F401 is silenced for
+# this re-export hub in pyproject.toml.
+from ..backend import _get_jax
+from ._block import block_lms
+from ._kernels_jax import _JITTED_EQ
+from ._kernels_numba import _get_numba
+from .blind import block_cma, block_rde, build_pilot_ref
+from .linear import apply_taps, estimate_transfer_function, zf_equalizer
+from .polarization import (
+    apply_interpolated_matrix,
+    demultiplex_polarization_tones_dynamic,
+    demultiplex_polarization_tones_static,
+)
+from .result import CPRState, EqualizerResult
+from .sequential import _check_rls_divergence, cma, lms, rde, rls
+
+__all__ = [
+    "CPRState",
+    "EqualizerResult",
+    "apply_interpolated_matrix",
+    "apply_taps",
+    "block_cma",
+    "block_lms",
+    "block_rde",
+    "build_pilot_ref",
+    "cma",
+    "demultiplex_polarization_tones_dynamic",
+    "demultiplex_polarization_tones_static",
+    "estimate_transfer_function",
+    "lms",
+    "rde",
+    "rls",
+    "zf_equalizer",
+]