r4178 - in /branches/N_state_model: generic_fns/align_tensor.py prompt/align_tensor.py -- December 11, 2007 - 14:35

Author: bugman
Date: Tue Dec 11 14:35:37 2007
New Revision: 4178

URL: http://svn.gna.org/viewcvs/relax?rev=4178&view=rev
Log:
Started to implement the generic_fns.align_tensor.singular_values() function.


Modified:
    branches/N_state_model/generic_fns/align_tensor.py
    branches/N_state_model/prompt/align_tensor.py

Modified: branches/N_state_model/generic_fns/align_tensor.py
URL: 
http://svn.gna.org/viewcvs/relax/branches/N_state_model/generic_fns/align_tensor.py?rev=4178&r1=4177&r2=4178&view=diff
==============================================================================
--- branches/N_state_model/generic_fns/align_tensor.py (original)
+++ branches/N_state_model/generic_fns/align_tensor.py Tue Dec 11 14:35:37 
2007
@@ -23,6 +23,7 @@
 # Python module imports.
 from copy import deepcopy
 from math import cos, pi, sin
+from numpy import float64, linalg, transpose, zeros
 from re import search
 
 # relax module imports.
@@ -754,3 +755,47 @@
 
     if orient_params:
         fold_angles()
+
+
+def singular_values():
+    """Function for calculating the singular values of all the loaded 
tensors.
+
+    The matrix on which SVD will be performed is:
+
+        | Sxx1 Syy1 Sxy1 Sxz1 Syz1 |
+        | Sxx2 Syy2 Sxy2 Sxz2 Syz2 |
+        | Sxx3 Syy3 Sxy3 Sxz3 Syz3 |
+        |  .    .    .    .    .   |
+        |  .    .    .    .    .   |
+        |  .    .    .    .    .   |
+        | SxxN SyyN SxyN SxzN SyzN |
+    """
+
+    # Alias the current data pipe.
+    cdp = relax_data_store[relax_data_store.current_pipe]
+
+    # Create the matrix to apply SVD on.
+    matrix = zeros((len(cdp.align_tensor), 5), float64)
+
+    # Pack the elements.
+    i = 0
+    for key in cdp.align_tensor.keys():
+        # The elements.
+        matrix[i,0] = cdp.align_tensor[key].Azz
+        matrix[i,1] = cdp.align_tensor[key].Axxyy
+        matrix[i,2] = cdp.align_tensor[key].Axy
+        matrix[i,3] = cdp.align_tensor[key].Axz
+        matrix[i,4] = cdp.align_tensor[key].Ayz
+
+        # Increment the index.
+        i = i + 1
+
+    print `matrix`
+
+    # SVD.
+    u, s, vh = linalg.svd(matrix)
+
+    # Store the singular values.
+    #cdp.align_tensor.singular_vals = s
+    print `s`
+

Modified: branches/N_state_model/prompt/align_tensor.py
URL: 
http://svn.gna.org/viewcvs/relax/branches/N_state_model/prompt/align_tensor.py?rev=4178&r1=4177&r2=4178&view=diff
==============================================================================
--- branches/N_state_model/prompt/align_tensor.py (original)
+++ branches/N_state_model/prompt/align_tensor.py Tue Dec 11 14:35:37 2007
@@ -286,4 +286,4 @@
             print text
 
         # Execute the functional code.
-        align_tensor.singular_vals()
+        align_tensor.singular_values()