Please see the manual in ../Docs

CAVEATS

* The specification of the 3D box for plotting is limited. Only a
  "base plane" (originally implemented for the 2D capability) can be
  fully specified with the options given in the manual. Within it, the
  a and b axes are chosen to be perpendicular to each other. The third
  dimension of the box is forced to be perpendicular to the
  plane. Hence, only orthorhombic boxes are allowed.

  Notably for wf plotting, it might be advantageous to provide an
  option for using the *original cell* as 3D box, or an arbitrary box.

* Many files can be produced for each wavefunction. It might be more
  economical to generate a single 'gridfunc' file with an extra
  dimension holding the (up to four, for spinors) components of the
  wavefunction. Then, a special version of the 'grid2cube' program
  could be generate a cube file with the desired incantation of the
  data (real or imaginary part, phase, modulus, etc).

* Atoms are not correctly replicated to the whole extent of the
  plotting box. This might be fixed automatically by the above
  two-step generation of cube files, as the "new generation" grid2cube
  program (g2c_ng) already treats atoms correctly.
  



