Package maths_fns :: Module pcs
[hide private]
[frames] | no frames]

Source Code for Module maths_fns.pcs

  1  ############################################################################### 
  2  #                                                                             # 
  3  # Copyright (C) 2008-2012 Edward d'Auvergne                                   # 
  4  #                                                                             # 
  5  # This file is part of the program relax (http://www.nmr-relax.com).          # 
  6  #                                                                             # 
  7  # This program is free software: you can redistribute it and/or modify        # 
  8  # it under the terms of the GNU General Public License as published by        # 
  9  # the Free Software Foundation, either version 3 of the License, or           # 
 10  # (at your option) any later version.                                         # 
 11  #                                                                             # 
 12  # This program is distributed in the hope that it will be useful,             # 
 13  # but WITHOUT ANY WARRANTY; without even the implied warranty of              # 
 14  # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the               # 
 15  # GNU General Public License for more details.                                # 
 16  #                                                                             # 
 17  # You should have received a copy of the GNU General Public License           # 
 18  # along with this program.  If not, see <http://www.gnu.org/licenses/>.       # 
 19  #                                                                             # 
 20  ############################################################################### 
 21   
 22  # Module docstring. 
 23  """Module for the calculation of pseudocontact shifts.""" 
 24   
 25  # Python imports. 
 26  from numpy import dot, sum 
 27   
 28   
29 -def ave_pcs_tensor(dj, vect, N, A, weights=None):
30 """Calculate the ensemble average PCS, using the 3D tensor. 31 32 This function calculates the average PCS for a set of XH bond vectors from a structural ensemble, using the 3D tensorial form of the alignment tensor. The formula for this ensemble average PCS value is:: 33 34 _N_ 35 \ T 36 <delta_ij(theta)> = > pc . djc . mu_jc . Ai . mu_jc, 37 /__ 38 c=1 39 40 where: 41 - i is the alignment tensor index, 42 - j is the index over spins, 43 - c is the index over the states or multiple structures, 44 - N is the total number of states or structures, 45 - theta is the parameter vector, 46 - djc is the PCS constant for spin j and state c, 47 - pc is the population probability or weight associated with state c (equally weighted to 1/N if weights are not provided), 48 - mu_jc is the unit vector corresponding to spin j and state c, 49 - Ai is the alignment tensor. 50 51 The PCS constant is defined as:: 52 53 mu0 15kT 1 54 dj = --- ----- ---- , 55 4pi Bo**2 r**3 56 57 where: 58 - mu0 is the permeability of free space, 59 - k is Boltzmann's constant, 60 - T is the absolute temperature, 61 - Bo is the magnetic field strength, 62 - r is the distance between the paramagnetic centre (electron spin) and the nuclear spin. 63 64 65 @param dj: The PCS constants for each structure c for spin j. This should be an array with indices corresponding to c. 66 @type dj: numpy rank-1 array 67 @param vect: The electron-nuclear unit vector matrix. The first dimension corresponds to the structural index, the second dimension is the coordinates of the unit vector. The vectors should be parallel to the vector connecting the paramagnetic centre to the nuclear spin. 68 @type vect: numpy matrix 69 @param N: The total number of structures. 70 @type N: int 71 @param A: The alignment tensor. 72 @type A: numpy rank-2 3D tensor 73 @param weights: The weights for each member of the ensemble (the last member need not be supplied). 74 @type weights: numpy rank-1 array 75 @return: The average PCS value. 76 @rtype: float 77 """ 78 79 # Initial back-calculated PCS value. 80 val = 0.0 81 82 # No weights given. 83 if weights == None: 84 pc = 1.0 / N 85 weights = [pc] * N 86 87 # Missing last weight. 88 if len(weights) < N: 89 pN = 1.0 - sum(weights, axis=0) 90 weights = weights.tolist() 91 weights.append(pN) 92 93 # Back-calculate the PCS. 94 for c in range(N): 95 val = val + weights[c] * dj[c] * dot(vect[c], dot(A, vect[c])) 96 97 # Return the average PCS. 98 return val
99 100
101 -def ave_pcs_tensor_ddeltaij_dAmn(dj, vect, N, dAi_dAmn, weights=None):
102 """Calculate the ensemble average PCS gradient element for Amn, using the 3D tensor. 103 104 This function calculates the average PCS gradient for a set of electron-nuclear spin unit vectors (paramagnetic to the nuclear spin) from a structural ensemble, using the 3D tensorial form of the alignment tensor. The formula for this ensemble average PCS gradient element is:: 105 106 _N_ 107 ddelta_ij(theta) \ T dAi 108 ---------------- = > pc . djc . mu_jc . ---- . mu_jc, 109 dAmn /__ dAmn 110 c=1 111 112 where: 113 - i is the alignment tensor index, 114 - j is the index over spins, 115 - m, the index over the first dimension of the alignment tensor m = {x, y, z}. 116 - n, the index over the second dimension of the alignment tensor n = {x, y, z}, 117 - c is the index over the states or multiple structures, 118 - theta is the parameter vector, 119 - Amn is the matrix element of the alignment tensor, 120 - djc is the PCS constant for spin j and state c, 121 - N is the total number of states or structures, 122 - pc is the population probability or weight associated with state c (equally weighted to 1/N if weights are not provided), 123 - mu_jc is the unit vector corresponding to spin j and state c, 124 - dAi/dAmn is the partial derivative of the alignment tensor with respect to element Amn. 125 126 127 @param dj: The PCS constants for each structure c for spin j. This should be an array with indices corresponding to c. 128 @type dj: numpy rank-1 array 129 @param vect: The electron-nuclear unit vector matrix. The first dimension corresponds to the structural index, the second dimension is the coordinates of the unit vector. The vectors should be parallel to the vector connecting the paramagnetic centre to the nuclear spin. 130 @type vect: numpy matrix 131 @param N: The total number of structures. 132 @type N: int 133 @param dAi_dAmn: The alignment tensor derivative with respect to parameter Amn. 134 @type dAi_dAmn: numpy rank-2 3D tensor 135 @param weights: The weights for each member of the ensemble (the last member need not be supplied). 136 @type weights: numpy rank-1 array 137 @return: The average PCS gradient element. 138 @rtype: float 139 """ 140 141 # Initial back-calculated PCS gradient. 142 grad = 0.0 143 144 # No weights given. 145 if weights == None: 146 pc = 1.0 / N 147 weights = [pc] * N 148 149 # Missing last weight. 150 if len(weights) < N: 151 pN = 1.0 - sum(weights, axis=0) 152 weights = weights.tolist() 153 weights.append(pN) 154 155 # Back-calculate the PCS gradient element. 156 for c in range(N): 157 grad = grad + weights[c] * dj[c] * dot(vect[c], dot(dAi_dAmn, vect[c])) 158 159 # Return the average PCS gradient element. 160 return grad
161 162
163 -def pcs_tensor(dj, mu, A):
164 """Calculate the PCS, using the 3D alignment tensor. 165 166 The PCS value is:: 167 168 T 169 delta_ij(theta) = dj . mu_j . Ai . mu_j, 170 171 where: 172 - i is the alignment tensor index, 173 - j is the index over spins, 174 - theta is the parameter vector, 175 - dj is the PCS constant for spin j, 176 - mu_j i the unit vector corresponding to spin j, 177 - Ai is the alignment tensor. 178 179 The PCS constant is defined as:: 180 181 mu0 15kT 1 182 dj = --- ----- ---- , 183 4pi Bo**2 r**3 184 185 where: 186 - mu0 is the permeability of free space, 187 - k is Boltzmann's constant, 188 - T is the absolute temperature, 189 - Bo is the magnetic field strength, 190 - r is the distance between the paramagnetic centre (electron spin) and the nuclear spin. 191 192 193 @param dj: The PCS constant for spin j. 194 @type dj: float 195 @param mu: The unit vector connecting the electron and nuclear spins. 196 @type mu: numpy rank-1 3D array 197 @param A: The alignment tensor. 198 @type A: numpy rank-2 3D tensor 199 @return: The PCS value. 200 @rtype: float 201 """ 202 203 # Return the PCS. 204 return dj * dot(mu, dot(A, mu))
205