The story of rubyist struggling with python :: Dict data with pycall

The story of rubyist struggling with python :: Dict data with pycall

PyCall Ruby version Tips has an almost perfect correspondence report. In addition, numpy-specific notation

{.example}

#Access to the last line m in python[-1,:]Where to write

And so on, rubyist even has an easy-to-understand explanation of numpy.

It looks like it can be done easily. .. ..

I'm sorry if it is a complicated data structure

However, when dealing with complex python data structures It's very difficult. The cause is that I don't know the key of dict. It's good when you take simple characters, but it doesn't seem to work well with slightly complicated characters.

Vasprun :: tdos or in pymatgen as follows Vasprun :: compolete \ _dos etc. can be taken.

However,

If p spd \ _dos = dosrun.complete \ _dos.get \ _spd \ _dos ()

{<OrbitalType.s: 0>: <pymatgen.electronic_structure.dos.Dos object at 0x102dd4710>, <OrbitalType.p: 1>: <pymatgen.electronic_structure.dos.Dos object at 0x102dc2dd8>, <OrbitalType.d: 2>: <pymatgen.electronic_structure.dos.Dos object at 0x102d962b0>}

Is output. I think this can be taken with spd \ _dos ('OrbitalType.s') No good. Therefore,

p s_d = spd_dos.to_a[0][1].densities

As

{.example}

array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00,
         0.00000000e+00,   0.00000000e+00,   0.00000000e+00,
...
      ])

Make it python :: array and then

s_list = PyCall::List.(s_d.to_a[0][1]).to_a

As an array, hash is taken out, and vals is taken out, and it is converted into a data structure that can be interpreted by ruby with to \ _a.

Once you get used to it, you won't have much trouble because the same code is repeated, Maybe I'll forget it over time. I want you to incorporate it in lib, but it seems that only characters read dict directly. .. .. Would you like to write a code and try pull req?

Densities and energy can't be helped without knowing the name, You can find this by reading the source code of pymatgen. Apparently, to use pycall, you need to repeat this.

{.ruby}

require 'pycall/import'
include PyCall::Import

pyimport :sys
pyimport 'numpy', as: :np
pyfrom 'pymatgen.io.vasp.outputs', import: :Vasprun

# dos
dosrun = Vasprun.("../dos_calc_fine/vasprun.xml")
e_object = (dosrun.tdos.energies - dosrun.efermi)
energy = PyCall::List.(e_object).to_a
spd_dos = dosrun.complete_dos.get_spd_dos()
spd = PyCall::Dict.(spd_dos)
s_d = spd.to_a[0][1].densities
s_list = PyCall::List.(s_d.to_a[0][1]).to_a
p_d = spd.to_a[1][1].densities
p_list = PyCall::List.(p_d.to_a[0][1]).to_a
d_d = spd.to_a[2][1].densities
d_list = PyCall::List.(d_d.to_a[0][1]).to_a
tdos = dosrun.tdos.densities
t_list = PyCall::List.(tdos.to_a[0][1]).to_a

require 'numo/gnuplot'
Numo.gnuplot do
  set yrange: '[-10:10]'
  plot [s_list, energy, w: :l, title: 's'],
  [p_list, energy, w: :l, title: 'p'],
  [d_list, energy, w: :l, title: 'd'],
  [t_list, energy, w: :l, title: 'total']
end