better use of templates for balance classes * aepbalance.h, aepbalance.cc: New files generated from base-aepbal.h, CmplxAEPBAL.cc, CmplxAEPBAL.h, dbleAEPBAL.cc, dbleAEPBAL.h, fCmplxAEPBAL.cc, fCmplxAEPBAL.h, floatAEPBAL.cc, and floatAEPBAL.h and making them templates. * gepbalance.h, gepbalance.cc: New files generate from CmplxGEPBAL.cc, CmplxGEPBAL.h, dbleGEPBAL.cc, dbleGEPBAL.h, fCmplxGEPBAL.cc, fCmplxGEPBAL.h, floatGEPBAL.cc, and floatGEPBAL.h and making them templates. * liboctave/numeric/module.mk: Update. * balance.cc, CMatrix.h, dMatrix.h, fCMatrix.h, fMatrix.h, mx-defs.h, mx-ext.h: Use new classes.

--- a/liboctave/array/dMatrix.h	Mon Feb 15 22:33:45 2016 -0500
+++ b/liboctave/array/dMatrix.h	Tue Feb 16 00:32:29 2016 -0500
@@ -43,6 +43,12 @@
   typedef ColumnVector column_vector_type;
   typedef RowVector row_vector_type;
 
+  typedef ColumnVector real_column_vector_type;
+  typedef RowVector real_row_vector_type;
+
+  typedef Matrix real_matrix_type;
+  typedef ComplexMatrix complex_matrix_type;
+
   typedef void (*solve_singularity_handler) (double rcon);
 
   Matrix (void) : NDArray () { }