-
Notifications
You must be signed in to change notification settings - Fork 1
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
PR: ADD YARP-XTB to PYTEST/Github-Workflow!
now you can run the whole xtb-level of YARP with Pytest/Github-Workflow (halfway testing)!
- Loading branch information
Showing
68 changed files
with
8,520 additions
and
2 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,16 @@ | ||
[orca] | ||
# ORCA needs the full path to its binary, so please provide the full path. | ||
#cmd=/sw/pkgs/lsa/orca/5.0.3/bin/orca | ||
|
||
[psi4] | ||
# As the Psi4 installation without conda is, to put it slightly, tricky it was | ||
# decided to allow the installation of Psi4 into a separate conda environment. | ||
# pysisyphus then creates a Psi4 input and sends it to the (bash)-script given below | ||
# that accepts/expects one argument. It is the responsibility of the scrip to activate | ||
# the appropriate conda environment and submit the Psi4 input. An example runpsi4.sh | ||
# script is given below. | ||
#cmd=/user/johannes/bin/runpsi4.sh | ||
|
||
[xtb] | ||
# Cmd to execute. Please ensure that xtb is on your $PATH. | ||
cmd=CONDA_ENV/bin/xtb |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,40 @@ | ||
from typing import Dict, List | ||
import torch | ||
import torch.nn as nn | ||
from torch import Tensor | ||
|
||
|
||
elem_to_num = {'H':1, 'C':6, 'N':7, 'O':8} | ||
|
||
class EnsembledModel(nn.Module): | ||
def __init__(self, models: List, x=['coord', 'numbers', 'charge'], out=['energy'], detach=False): | ||
super().__init__() | ||
self.models = nn.ModuleList(models) | ||
self.x = x | ||
self.out = out | ||
self.detach = detach | ||
|
||
def forward(self, data: Dict[str, Tensor]) -> Dict[str, Tensor]: | ||
res : List[Dict[str, Tensor]] = [] | ||
for model in self.models: | ||
_in = dict() | ||
for k in data: | ||
if k in self.x: | ||
_in[k] = data[k] | ||
_out = model(_in) | ||
_r = dict() | ||
for k in _out: | ||
if k in self.out: | ||
_r[k] = _out[k] | ||
if self.detach: | ||
_r[k] = _r[k].detach() | ||
res.append(_r) | ||
|
||
for k in res[0]: | ||
v = [] | ||
for x in res: | ||
v.append(x[k]) | ||
vv = torch.stack(v, dim=0) | ||
data[k] = vv.mean(dim=0) | ||
data[k + '_std'] = vv.std(dim=0) | ||
return data |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,160 @@ | ||
#!/bin/env python | ||
# Author: Hsuan-Hao Hsu ([email protected]) | ||
import os,sys | ||
import numpy as np | ||
import yaml | ||
import logging | ||
import time | ||
import json | ||
import pickle | ||
import pyjokes | ||
import fnmatch | ||
from xgboost import XGBClassifier | ||
|
||
from yarp.input_parsers import xyz_parse | ||
from wrappers.orca import ORCA | ||
from wrappers.crest import CREST | ||
from utils import * | ||
from constants import Constants | ||
from job_submission import * | ||
from wrappers.gaussian import Gaussian | ||
from job_mapping import * | ||
|
||
# This program aims to refine the TSs (.xyz files) by DFT level. | ||
# We don't have any info of reactant and product. | ||
# Just do TS-opt and IRC calculations. | ||
def main(args): | ||
TS_dict=dict() | ||
# read TS into dictionary | ||
if os.path.isfile(args["input"]): | ||
E, G=xyz_parse(args["input"]) | ||
TS_dict[args["input"].split("/")[-1].split(".")[0]]=dict() | ||
TS_dict[args["input"].split("/")[-1].split(".")[0]]["E"]=E | ||
TS_dict[args["input"].split("/")[-1].split(".")[0]]["TSG"]=G | ||
else: | ||
xyz_files=[args["input"]+"/"+i for i in os.listdir(args["input"]) if fnmatch.fnmatch(i, "*.xyz")] | ||
for i in xyz_files: | ||
E, G=xyz_parse(i) | ||
TS_dict[i.split("/")[-1].split(".")[0]]=dict() | ||
TS_dict[i.split("/")[-1].split(".")[0]]["E"]=E | ||
TS_dict[i.split("/")[-1].split(".")[0]]["TSG"]=G | ||
# finish laod initial TSs into a dict | ||
scratch=args["scratch"] | ||
if os.path.isdir(args["scratch"]) is False: os.mkdir(args["scratch"]) | ||
if len(args["dft_lot"].split()) > 1: dft_lot="/".join(args["dft_lot"].split()) | ||
else: dft_lot=args["dft_lot"] | ||
# run TS optimization | ||
job_list=dict() | ||
running_jobs=[] | ||
for i in TS_dict.keys(): | ||
wf=f"{scratch}/{i}" | ||
if os.path.isdir(wf) is False: os.mkdir(wf) | ||
xyz_file=f"{wf}/{i}.xyz" | ||
xyz_write(xyz_file, TS_dict[i]["E"], TS_dict[i]["TSG"]) | ||
if args["package"]=="ORCA": | ||
dft_job=ORCA(input_geo=xyz_file, work_folder=wf, nproc=int(args["dft_nprocs"]), mem=int(args["mem"])*1000, jobname=f"{i}-TSOPT",\ | ||
jobtype="OptTS Freq", lot=args["dft_lot"], charge=args["charge"], multiplicity=args["multiplicity"], solvent=args["solvent"],\ | ||
solvation_model=args["solvation_model"], dielectric=args["dielectric"], writedown_xyz=True) | ||
dft_job.generate_geometry_settings(hess=True, hess_step=int(args["hess_recalc"])) | ||
dft_job.generate_input() | ||
job_list[i]=dft_job | ||
if dft_job.calculation_terminated_normally() is False: running_jobs.append(i) | ||
elif args["package"]=="Gaussian": | ||
dft_job=Gaussian(input_geo=xyz_file, work_folder=wf, nproc=int(args["dft_nprocs"]), mem=int(args["mem"])*1000, jobname=f"{i}-TSOPT",\ | ||
jobtype="tsopt", lot=dft_lot, charge=args["charge"], multiplicity=args["multiplicity"], solvent=args["solvent"],\ | ||
solvation_model=args["solvation_model"], dielectric=args["dielectric"], dispersion=args["dispersion"]) | ||
dft_job.generate_input() | ||
job_list[i]=dft_job | ||
if dft_job.calculation_terminated_normally() is False: running_jobs.append(i) | ||
if len(running_jobs)>1: | ||
n_submit=len(running_jobs)//int(args["dft_njobs"]) | ||
if len(running_jobs)%int(args["dft_njobs"])>0: n_submit+=1 | ||
startid=0 | ||
slurm_jobs=[] | ||
startid=0 | ||
for i in range(n_submit): | ||
slurmjob=SLURM_Job(jobname=f"TSOPT.{i}", ppn=int(args["ppn"]), partition=args["partition"], time=args["dft_wt"], mem_per_cpu=int(args["mem"]*1100)) | ||
endid=min(startid+int(args["dft_njobs"]), len(running_jobs)) | ||
if args["package"]=="ORCA": slurmjob.create_orca_jobs([job_list[ind] for ind in running_jobs[startid:endid]]) | ||
elif args["package"]=="Gaussian": slurmjob.create_gaussian_jobs([job_list[ind] for ind in running_jobs[startid:endid]]) | ||
slurmjob.submit() | ||
startid=endid | ||
slurm_jobs.append(slurmjob) | ||
print(f"Running {len(slurm_jobs)} ts optimization jobs...") | ||
monitor_jobs(slurm_jobs) | ||
key=[i for i in job_list.keys()] | ||
for i in key: | ||
dft_opt=job_list[i] | ||
if dft_opt.calculation_terminated_normally() and dft_opt.optimization_converged() and dft_opt.is_TS(): | ||
_, geo=dft_opt.get_final_structure() | ||
if dft_lot not in TS_dict[i].keys(): TS_dict[i][dft_lot]=dict() | ||
TS_dict[i][dft_lot]["geo"]=geo | ||
TS_dict[i][dft_lot]["thermal"]=dft_opt.get_thermal() | ||
#TS_dict[i][dft_lot]["SPE"]=dft_opt.get_energy() | ||
TS_dict[i][dft_lot]["imag_mode"]=dft_opt.get_imag_freq_mode() | ||
else: | ||
print("No ts optimiation jobs need to be performed...") | ||
|
||
# Finish running TS-opt jobs | ||
# Prepare IRC jobs | ||
job_list=dict() | ||
running_jobs=[] | ||
for i in TS_dict.keys(): | ||
wf=f"{scratch}/{i}" | ||
xyz_file=f"{wf}/{i}.xyz" | ||
if dft_lot not in TS_dict[i].keys(): continue | ||
xyz_write(xyz_file, TS_dict[i]["E"], TS_dict[i][dft_lot]["geo"]) | ||
if args["package"]=="ORCA": | ||
dft_job=ORCA(input_geo=xyz_file, work_folder=wf, nproc=int(args["dft_nprocs"]), mem=int(args["mem"])*1000, jobname=f"{i}-IRC",\ | ||
jobtype="IRC", lot=args["dft_lot"], charge=args["charge"], multiplicity=args["multiplicity"], solvent=args["solvent"],\ | ||
solvation_model=args["solvation_model"], dielectric=args["dielectric"], writedown_xyz=True) | ||
dft_job.generate_geometry_settings(hess=True, hess_step=int(args["hess_recalc"])) | ||
dft_job.generate_input() | ||
job_list[i]=dft_job | ||
if dft_job.calculation_terminated_normally() is False: running_jobs.append(i) | ||
elif args["package"]=="Gaussian": | ||
dft_job=Gaussian(input_geo=xyz_file, work_folder=wf, nproc=int(args["dft_nprocs"]), mem=int(args["mem"])*1000, jobname=f"{i}-IRC",\ | ||
jobtype="irc", lot=dft_lot, charge=args["charge"], multiplicity=args["multiplicity"], solvent=args["solvent"],\ | ||
solvation_model=args["solvation_model"], dielectric=args["dielectric"], dispersion=args["dispersion"]) | ||
dft_job.generate_input() | ||
job_list[i]=dft_job | ||
if dft_job.calculation_terminated_normally() is False: running_jobs.append(i) | ||
if len(running_jobs)>1: | ||
n_submit=len(running_jobs)//int(args["dft_njobs"]) | ||
if len(running_jobs)%int(args["dft_njobs"])>0: n_submit+=1 | ||
startid=0 | ||
slurm_jobs=[] | ||
for i in range(n_submit): | ||
slurmjob=SLURM_Job(jobname=f"IRC.{i}", ppn=int(args["ppn"]), partition=args["partition"], time=args["dft_wt"], mem_per_cpu=int(args["mem"]*1100)) | ||
endid=min(startid+int(args["dft_njobs"]), len(running_jobs)) | ||
if args["package"]=="ORCA": slurmjob.create_orca_jobs([job_list[ind] for ind in running_jobs[startid:endid]]) | ||
elif args["package"]=="Gaussian": slurmjob.create_gaussian_jobs([job_list[ind] for ind in running_jobs[startid:endid]]) | ||
slurmjob.submit() | ||
startid=endid | ||
slurm_jobs.append(slurmjob) | ||
print(f"Running {len(slurm_jobs)} irc jobs...") | ||
monitor_jobs(slurm_jobs) | ||
key=[i for i in job_list.keys()] | ||
for i in key: | ||
dft_opt=job_list[i] | ||
if dft_opt.calculation_terminated_normally(): | ||
job_success=False | ||
try: | ||
E, G1, G2, TSG, barrier1, barrier2=dft_opt.analyze_IRC() | ||
job_success=True | ||
except: pass | ||
if job_success==True: | ||
TS_dict[i][dft_lot]["IRC"]=dict() | ||
TS_dict[i][dft_lot]["IRC"]["node"]=[G1, G2] | ||
TS_dict[i][dft_lot]["IRC"]["TS"]=TSG | ||
TS_dict[i][dft_lot]["barriers"]=[barrier2, barrier1] | ||
else: | ||
print("No irc jobs need to be performed...") | ||
with open(args["reaction_data"], 'wb') as f: | ||
pickle.dump(TS_dict, f) | ||
return | ||
|
||
if __name__=="__main__": | ||
parameters = sys.argv[1] | ||
parameters = yaml.load(open(parameters, "r"), Loader=yaml.FullLoader) | ||
main(parameters) |
Oops, something went wrong.