# ------------------------------------------------------------------------ #
# Copyright 2022 SPTK Working Group #
# #
# Licensed under the Apache License, Version 2.0 (the "License"); #
# you may not use this file except in compliance with the License. #
# You may obtain a copy of the License at #
# #
# http://www.apache.org/licenses/LICENSE-2.0 #
# #
# Unless required by applicable law or agreed to in writing, software #
# distributed under the License is distributed on an "AS IS" BASIS, #
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. #
# See the License for the specific language governing permissions and #
# limitations under the License. #
# ------------------------------------------------------------------------ #
import warnings
import torch
from ..utils.private import check_size
from ..utils.private import get_values
from .base import BaseFunctionalModule
from .lpc2par import LinearPredictiveCoefficientsToParcorCoefficients
from .par2lpc import ParcorCoefficientsToLinearPredictiveCoefficients
[docs]
class LinearPredictiveCoefficientsStabilityCheck(BaseFunctionalModule):
"""See `this page <https://sp-nitech.github.io/sptk/latest/main/lpccheck.html>`_
for details.
Parameters
----------
lpc_order : int >= 0
The order of the LPC, :math:`M`.
margin : float in (0, 1)
The margin to guarantee the stability of LPC.
warn_type : ['ignore', 'warn', 'exit']
The warning type.
"""
def __init__(self, lpc_order, margin=1e-16, warn_type="warn"):
super().__init__()
self.in_dim = lpc_order + 1
self.values = self._precompute(*get_values(locals()))
[docs]
def forward(self, a):
"""Check the stability of the input LPC coefficients.
Parameters
----------
a : Tensor [shape=(..., M+1)]
The input LPC coefficients.
Returns
-------
out : Tensor [shape=(..., M+1)]
The modified LPC coefficients.
Examples
--------
>>> x = diffsptk.nrand(4)
tensor([-0.9966, -0.2970, -0.2173, 0.0594, 0.5831])
>>> lpc = diffsptk.LPC(3, 5)
>>> a = lpc(x)
>>> a
tensor([ 1.1528, -0.2613, -0.0274, 0.1778])
>>> lpccheck = diffsptk.LinearPredictiveCoefficientsStabilityCheck(3)
>>> a2 = lpccheck(a)
tensor([ 1.1528, -0.2613, -0.0274, 0.1778])
"""
check_size(a.size(-1), self.in_dim, "dimension of LPC")
return self._forward(a, *self.values)
@staticmethod
def _func(a, *args, **kwargs):
values = LinearPredictiveCoefficientsStabilityCheck._precompute(
a.size(-1) - 1, *args, **kwargs
)
return LinearPredictiveCoefficientsStabilityCheck._forward(a, *values)
@staticmethod
def _takes_input_size():
return True
@staticmethod
def _check(lpc_order, margin):
if lpc_order < 0:
raise ValueError("lpc_order must be non-negative.")
if not 0 < margin < 1:
raise ValueError("margin must be in (0, 1).")
@staticmethod
def _precompute(lpc_order, margin, warn_type):
LinearPredictiveCoefficientsStabilityCheck._check(lpc_order, margin)
return (1 - margin, warn_type)
@staticmethod
def _forward(a, bound, warn_type):
k = LinearPredictiveCoefficientsToParcorCoefficients._func(a)
K, k1 = torch.split(k, [1, k.size(-1) - 1], dim=-1)
if torch.any(1 <= torch.abs(k1)):
if warn_type == "ignore":
pass
elif warn_type == "warn":
warnings.warn("Detected unstable LPC coefficients.")
elif warn_type == "exit":
raise RuntimeError("Detected unstable LPC coefficients.")
else:
raise RuntimeError
k1 = torch.clip(k1, -bound, bound)
k2 = torch.cat((K, k1), dim=-1)
a2 = ParcorCoefficientsToLinearPredictiveCoefficients._func(k2)
return a2