extracting forces and energies for lammps data

[nisargjoshi10]

Hi all,

I had a question about extracting forces and energies from lammps simulation data for dpata. I am able to get the coords.raw, box.raw and type.raw from the trajectory dump file. My question is if there's any specific format dpdata requires for forces and energies. I have been dumping the fx, fy and fz forces in a dump file but when I try to load that using dpdata I get this error:

ValueError: 'xs' is not in list

Same with energy, I have energy output in log.lammps and out.lammps but not sure how to extract that with dpdata either. Would appreciate your help in extracting all the information for creating the raw files.

Thanks,
Nisarg

[wanghan-iapcm]

Currently dpdata does not support extracting labels from lammps output. You are welcome to contribute.

[nisargjoshi10]

thanks for your response! is there still a way to extract the forces from the dump file?

[wanghan-iapcm]

One may implement the reading of forces similar to the reading of coordinates.

dpdata/dpdata/lammps/dump.py

Lines 54 to 85 in 24ff734

	def get_coordtype_and_scalefactor(keys):
	# 4 types in total,with different scaling factor
	key_pc=['x','y','z'] # plain cartesian, sf = 1
	key_uc=['xu','yu','zu'] # unwraped cartesian, sf = 1
	key_s=['xs','ys','zs'] # scaled by lattice parameter, sf = lattice parameter
	key_su = ['xsu','ysu','zsu'] #scaled and unfolded,sf = lattice parameter
	lmp_coor_type = [key_pc,key_uc,key_s,key_su]
	sf = [0,0,1,1]
	for k in range(4):
	if all(i in keys for i in lmp_coor_type[k]):
	return lmp_coor_type[k],sf[k]
	def safe_get_posi(lines,cell,orig=np.zeros(3)) :
	blk, head = _get_block(lines, 'ATOMS')
	keys = head.split()
	coord_tp_and_sf = get_coordtype_and_scalefactor(keys)
	assert coord_tp_and_sf is not None, 'Dump file does not contain atomic coordinates!'
	coordtype, sf = coord_tp_and_sf
	id_idx = keys.index('id') - 2
	xidx = keys.index(coordtype[0])-2
	yidx = keys.index(coordtype[1])-2
	zidx = keys.index(coordtype[2])-2
	sel = (xidx, yidx, zidx)
	posis = []
	for ii in blk :
	words = ii.split()
	posis.append([float(words[id_idx]), float(words[xidx]), float(words[yidx]), float(words[zidx])])
	posis.sort()
	posis = np.array(posis)[:,1:4]
	if not sf:
	posis = (posis-orig)@np.linalg.inv(cell)# Convert to scaled coordinates for unscaled coordinates
	return (posis%1)@cell # Convert scaled coordinates back to Cartesien coordinates

[nisargjoshi10]

yeah that's what I've been trying, writing the force coordinates to a dump file and then reading them through dpdata. But while doing that I'm getting this error and I can't load the dump file

ValueError: 'xs' is not in list

I'll try to look into the code and see if I can make it work somehow!

[nisargjoshi10]

this is the output from my code:

>>> d_force = dpdata.System('force_wte.dump')
Traceback (most recent call last):
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/lammps/dump.py", line 88, in safe_get_posi
    posis = get_posi(lines) - orig
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/lammps/dump.py", line 58, in get_posi
    xidx = keys.index('x') - 2
ValueError: 'x' is not in list

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/system.py", line 129, in __init__
    self.from_fmt(file_name, fmt, type_map=type_map, begin= begin, step=step, **kwargs)
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/system.py", line 146, in from_fmt
    func(self, file_name, **kwargs)
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/system.py", line 567, in from_lammps_dump
    self.data = dpdata.lammps.dump.system_data(lines, type_map)
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/lammps/dump.py", line 176, in system_data
    system['coords'] = [safe_get_posi(lines, cell, np.array(orig))]
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/lammps/dump.py", line 90, in safe_get_posi
    fposis = get_posi_frac(lines)
  File "/gscratch/pfaendtner/nisargj/codes/miniconda/envs/deepmd/lib/python3.9/site-packages/dpdata/lammps/dump.py", line 74, in get_posi_frac
    xidx = keys.index('xs') - 2
ValueError: 'xs' is not in list```

[njzjz]

So, if I understand correctly, you only dump forces but no positions?

[nisargjoshi10]

I'm dumping forces and positions into separate dump files. Are they supposed to be in the same file?

[wanghan-iapcm]

Yes, the positions are supposed to always present in the dump file

[nisargjoshi10]

okay I'll try adding both in the same dump file and check. Thanks for your help!