Simplified norm calculation in Rodrigues formulae. Minor grammar corrections in examples

Author: davidcorteso

examples/neb_micromagnetic/elongated_particle/elongated_particle_along_z.py

@@ -5,9 +5,9 @@
 
 Test of the NEBM on an elongated particle of cylindrical shape, based on [1],
 where a series of simple magnetic systems are tested using the NEBM in
-Spherical coordinates with an Euclidean distance. For the alongated particle
+Spherical coordinates with an Euclidean distance. For the elongated particle
 test, the system is a 70 nm long and 12 nm wide (originally it is 13 nm)
-nanocylinder, with an uniaxial anisotropy in the direction of the long axis.The
+nanocylinder, with a uniaxial anisotropy in the direction of the long axis.The
 test goal is to find the minimum energy path (MEP) between two degenerate
 ferromagnetic states (saturated states along the anisotropy axis), which are
 the ground states of this system. The MEP is given by a transverse domain wall

fidimag/common/nebm_tools.py

@@ -120,7 +120,7 @@ def linear_interpolation_spherical(y_initial, y_final, n, pins=None):
         #  Just use half the length of y_initial
         pins = np.zeros(len(y_initial[::2]))
 
-    # Since we have a pin index per every PAIR of coordinates, we copy very
+    # Since we have a pin index per every PAIR of coordinates, we copy every
     # entry. For example: [1 0] --> [1 1 0 0]
     # and we change only unpinned spins (0)
     _filter = np.repeat(pins, 2) == 0
@@ -177,7 +177,7 @@ def interpolation_Rodrigues_rotation(y_initial, y_final, n, pins=None):
         # Only use 1/3 of the length of y_initial (1 pin per mesh/lattice site)
         pins = np.zeros(len(y_initial[::3]))
 
-    # Since we have a pin index per every TRIAD of coordinates, we copy very
+    # Since we have a pin index per every TRIAD of coordinates, we copy every
     # entry. For example: [1 0] --> [1 1 1 0 0 0]
     # and we change only unpinned spins (0)
     _filter = np.repeat(pins, 3) == 0
@@ -191,8 +191,8 @@ def interpolation_Rodrigues_rotation(y_initial, y_final, n, pins=None):
     # The cross products of corresponding spins in the initial and final images
     yi_cross_yf = np.cross(y_initial, y_final)
     # This should only be an array of ones:
-    yi_cross_yf_norm = np.apply_along_axis(lambda x: np.sqrt(np.sum(x ** 2)),
-                                           1, yi_cross_yf)
+    yi_cross_yf_norm = np.linalg.norm(yi_cross_yf, axis=1)
+
     # The rotation axis is just the normalised cross product defined before
     rot_axis = yi_cross_yf / yi_cross_yf_norm[:, np.newaxis]
     rot_axis = np.cross(rot_axis, y_initial)