Edit on GitHub

VASP file I/O

We have provided methods to load:

  • POSCAR
  • PROCAR
  • PROOUT
  • EIGENVAL

POSCAR

PROCAR

PROCAR for LDA

For normal LDA or GGA calculation, you can load PROCAR file using load_procar function without any keyword arguments. And there are three values are returned, representing metadata of projection characters, k-points and bands, respectively.

projection, kpoints, bands = load_procar("PROCAR")

PROCAR for LSDA

For LSDA or GGA+Spin calculation, you need set the keyword argument spin to be true. And the returned values are still divided into these three parts.

projection, kpoints, bands= load_procar("PROCAR"; spin = true)

PROCAR for non-collinear calculation

For non-collinear calculation, you need set the keyword argument noncollinear to be true. The returned values are divided into five parts, representing total projection characters, projection characters of x-axis, projection characters of y-axis, projection characters of z-axis, k-points and bands, respectively.

projection_all, projection_x, projection_y, projection_z, kpoints, bands =
    load_procar("PROCAR"; noncollinear = true)

Trick to distinguish between major spin and minor spin

In addition, non-collinear calculation in VASP doesn't distinguish between major and minor spin. But you can use a calculation trick to distinguish between spins. For example, you can set the spin quantization axis to be z axis using SAXIS tag. Then, the only nonzero projection character value is projection_all and projection_z. For negative projection character value, the spin is minor spin, otherwise the spin is major spin. And we have defined a method distinguish_spin to do this job.

projection, bands = distinguish_spin(projection_all, projection_z, bands)

The metadata of projection character and band of major spin and minor spin have been stored in projection and bands with ProjectionWithSpin and BandsWithSpin data type respectively.