update example

examples/active_learning_qe/AL.sh

@@ -14,20 +14,20 @@
 # source XXX
 
 echo start AL sample
-srun -n 8 abics_mlref input.toml >> abics_mlref.out
+srun -n 8 abics_mlref input-aenet.toml >> abics_mlref.out
 
 echo start parallel_run 1
 sh parallel_run.sh
 
 echo start AL final
-srun -n 8 abics_mlref input.toml >> abics_mlref.out
+srun -n 8 abics_mlref input-aenet.toml >> abics_mlref.out
 
 #train
 # module purge
 # module load intel_compiler/2020.2.254  
 # module load intel_mpi/2020.2.254
 
 echo start training
-abics_train input.toml >> abics_train.out
+abics_train input-aenet.toml >> abics_train.out
 
 echo Done

examples/active_learning_qe/MC.sh

@@ -11,6 +11,6 @@
 # source XXX
 
 #sleep 30
-srun -n 8 abics_sampling input.toml >> abics_sampling.out
+srun -n 8 abics_sampling input-aenet.toml >> abics_sampling.out
 
 echo Done

examples/active_learning_qe/calc_DOI.py

@@ -14,19 +14,8 @@
 # You should have received a copy of the GNU General Public License
 # along with this program. If not, see http://www.gnu.org/licenses/.
 
-from mpi4py import MPI
 from pymatgen.core import Structure
 from pymatgen.analysis.structure_matcher import StructureMatcher, FrameworkComparator
-import os, sys
-import numpy as np
-import scipy.constants as constants
-from abics.mc_mpi import RXParams
-
-throwout = 1
-tomlfile = sys.argv[1] if len(sys.argv) > 1 else "input.toml"
-rxparams = RXParams.from_toml(tomlfile)
-nreplicas = rxparams.nreplicas
-comm = MPI.COMM_WORLD
 
 spinel_struct = Structure.from_file("MgAl2O4.vasp")
 Asite_struct = spinel_struct.copy()
@@ -51,30 +40,3 @@ def calc_DOI(structure):
             x += 1
     x /= float(len(asites))
     return x
-
-
-myT = comm.Get_rank()
-kTs = np.load(open("./kTs.npy", "rb"))
-
-if myT < nreplicas:
-    Trank_hist = np.load(open(str(myT) + "/Trank_hist.npy", "rb"))
-    nstep = len(Trank_hist)
-    os.chdir("Tseparate/{}".format(myT))
-    DOI = []
-    for i in range(nstep):
-        DOI.append(calc_DOI(Structure.from_file("structure.{}.vasp".format(i))))
-    DOI = np.array(DOI)
-    np.savetxt("DOI.dat", DOI)
-    DOImean = DOI[throwout:].mean()
-    os.chdir("..")
-    myT_kelvin = kTs[myT] / constants.value(u"Boltzmann constant in eV/K")
-
-for i in range(len(kTs)):
-    comm.Barrier()
-    if myT != i:
-        continue
-    outfi = open("DOI_T.dat", "a")
-    outfi.write("{:6.1f} \t".format(myT_kelvin))
-    outfi.write("{}\n".format(DOImean))
-    outfi.close()
-

examples/active_learning_qe/input-aenet.toml

@@ -120,5 +120,8 @@ num = 8 #216 #8000
 
 
 [observer]
-reference_structure = "MgAl2O4.vasp"
-ignored_species = ["Al", "O"]
+[[observer.solver]]
+name = "DOI"
+type = "user"
+function = "calc_DOI.calc_DOI"
+perturb = 0.0

examples/active_learning_qe/input-mlip-3.toml

@@ -119,5 +119,8 @@ num = 8 #216 #8000
 
 
 [observer]
-reference_structure = "MgAl2O4.vasp"
-ignored_species = ["Al", "O"]
+[[observer.solver]]
+name = "DOI"
+type = "user"
+function = "calc_DOI.calc_DOI"
+perturb = 0.0

examples/active_learning_qe/input-nequip.toml

@@ -117,5 +117,8 @@ num = 8 #216 #8000
 
 
 [observer]
-reference_structure = "MgAl2O4.vasp"
-ignored_species = ["Al", "O"]
+[[observer.solver]]
+name = "DOI"
+type = "user"
+function = "calc_DOI.calc_DOI"
+perturb = 0.0