sponge.colvar.group 源代码

# Copyright 2021-2023 @ Shenzhen Bay Laboratory &
#                       Peking University &
#                       Huawei Technologies Co., Ltd
#
# This code is a part of MindSPONGE:
# MindSpore Simulation Package tOwards Next Generation molecular modelling.
#
# MindSPONGE is open-source software based on the AI-framework:
# MindSpore (https://www.mindspore.cn/)
#
# 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.
# ============================================================================
"""
Atom group
"""

from typing import Union, List, Tuple
import mindspore
from mindspore import ops
from mindspore.ops import functional as F
from mindspore import Tensor
from mindspore.nn import CellList

from .colvar import Colvar
from ..function import get_integer


[文档]class ColvarGroup(Colvar): r"""Concatenate a group of `Colvar` classes into one `Colvar` class Args: colvar (list or tuple): Array of `Colvar` to be concatenated. axis (int): Axis to be concatenated. NOTE: This refers to the axis of the output Tensor with the shape `(B, S_1, S_2, ..., S_n)`. Default: -1. use_pbc (bool): Whether to use periodic boundary condition. Default: ``None``. name (str): Name of the collective variables. Default: 'colvar_group'. Supported Platforms: ``Ascend`` ``GPU`` """ def __init__(self, colvar: Union[List[Colvar], Tuple[Colvar]], axis: int = -1, use_pbc: bool = None, name: str = 'colvar_group', ): super().__init__(name=name) if isinstance(colvar, Colvar): colvar = [colvar] elif not isinstance(colvar, (list, tuple)): raise TypeError(f'The type of "colvar" must be list of Colvar but got: {type(colvar)}') self.num_colvar = len(colvar) axis = get_integer(axis) if axis == 0: raise ValueError(f'The axis ({axis}) cannot be 0 (the dimension of batch size)!') shape = None dim = 0 periodic = () colvar_ = [] for i, cv in enumerate(colvar): shape_ = (1,) + cv.shape dim += shape_[axis] if axis == -1: shape_ = shape_[:-1] + (None,) else: shape_ = shape_[:axis] + (None,) + shape_[axis+1:] if i > 0 and shape_ != shape: raise ValueError(f'The shape of the No.{i} colvar {cv.shape} cannot be ' f'concatenate with the shape of the colvar group: {shape}') shape = shape_ if use_pbc is not None: cv.set_pbc(use_pbc) colvar_.append(cv) periodic += (F.expand_dims(cv.periodic, 0),) self.colvar: List[Colvar] = CellList(colvar_) if axis == -1: shape = shape[:-1] + (dim,) else: shape = shape[:axis] + (dim,) + shape[axis+1:] self._shape = shape[1:] self._ndim = len(self._shape) self.concat = ops.Concat(axis) periodic_int = [p.astype(mindspore.int32) for p in periodic] self._periodic = F.squeeze(self.concat(periodic_int), 0).astype(mindspore.bool_)
[文档] def set_pbc(self, use_pbc: bool): """set whether to use periodic boundary condition""" self._use_pbc = use_pbc self.get_vector.set_pbc(use_pbc) for i in range(self.num_colvar): self.colvar[i].set_pbc(use_pbc) return self
def construct(self, coordinate: Tensor, pbc_box: Tensor = None): r"""get position coordinates of colvar group Args: coordinate (Tensor): Tensor of shape `(B, A, D)`. Data type is float. B means Batchsize, i.e. number of walkers in simulation. A means Number of colvar in system. D means Dimension of the simulation system. Usually is 3. Position coordinate of colvar in system. pbc_box (Tensor): Tensor of shape `(B, D)`. Data type is float. Tensor of PBC box. Default: ``None``. Returns: position (Tensor): Tensor of shape `(B, S_1, S_2, ..., S_n)`. Data type is float. """ colvar = () for i in range(self.num_colvar): # (B, a_1'(i), a_2, ..., a_n, D) colvar += (self.colvar[i](coordinate, pbc_box),) # (B, a_1, a_2, ..., a_n, D) <- (B, a_1'(i), a_2, ..., a_n, D) colvar = self.concat(colvar) if self.do_reshape: new_shape = coordinate.shape[0] + self._shape colvar = F.reshape(colvar, new_shape) return colvar