No subject

=C2=A0 vnl_matrix&lt;double&gt; B =3D <br>=C2=A0=C2=A0=C2=A0 vnl_matrix_inv=
erse&lt;double&gt;( N_matrix.transpose() * N_matrix) * N_matrix.transpose()=
 * data_matrix;<br>
<br>is from <br>
<br>Insight/Code/Common/<div id=3D":5w">itkNonUniformBSpline.txx<br>(line 3=
13).<br><a href=3D"http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/Commo=
n/itkNonUniformBSpline.txx?revision=3D1.9&amp;root=3DInsight&amp;view=3Dmar=
kup" target=3D"_blank">http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/C=
ommon/itkNonUniformBSpline.txx?revision=3D1.9&amp;root=3DInsight&amp;view=
=3Dmarkup</a><br>

<br><br>It is trying to solve the linear system<br><br>=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 X=C2=A0 =3D N * B<br=
><br>by doing:<br><br>=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 B =3D ( N&=
#39; * N )^-1=C2=A0 *=C2=A0 N&#39;=C2=A0 * X<br><br><br>instead of invoking=
 one of the linear solvers<br>that we are carrying in ITK via VXL.<br>

<br><br>----<br><br><br>=C2=A0=C2=A0 Luis<br><br></div>

--000e0ce0f792df210404942b30ba--