Data visualisation

[kaist-amsg]

Data visualisation

[sgbaird]

Ok gotcha. You might have better luck looking at ADMM or Rosetta repos and asking there if you get stuck.

[sgbaird]

@WikGitHub Shot in the dark here. Did you try the following function:

Composition-Conditioned-Crystal-GAN/Composition_Conditioned_Crystal_GAN/view_atoms_mgmno.py

Lines 63 to 76 in 35cf6d0

	def view_atoms(image, view = True):
	x = image
	x = x.reshape(-1,3)
	l = x[0,:]*30
	a = x[1,:]*180
	cell = np.hstack((l,a))
	pos=x[2:,:]
	pos,n_mg, n_mn,n_o = remove_zero_padding(pos)
	scaled_pos = back_to_10_cell(pos,n_mg,n_mn,n_o)
	atoms = back_to_real_cell(scaled_pos, cell, n_mg,n_mn,n_o)
	atoms.set_pbc([1,1,1])
	if view:
	atoms.edit()
	return atoms, x

where you load the .npy file and pass that array as the image argument?

[sgbaird]

@WikGitHub ah, gotcha. I'm totally out of my element there. Did some simple Google searches but not much luck. Are you sure that the .npy files contain information related to protein structure? I figured Composition-Conditioned-Crystal-GAN was just for crystal structure. Maybe there's some way to infer protein structure from a collection of atoms and atom locations.