def test_runs_assimilate(self):
drone = VaspDrone(runs=["relax1", "relax2"])
doc = drone.assimilate(self.relax2)
oszicar2 = Oszicar(os.path.join(self.relax2, "OSZICAR.relax2.gz"))
outcar1 = Outcar(os.path.join(self.relax2, "OUTCAR.relax1.gz"))
outcar2 = Outcar(os.path.join(self.relax2, "OUTCAR.relax2.gz"))
outcar1 = outcar1.as_dict()
outcar2 = outcar2.as_dict()
run_stats1 = outcar1.pop("run_stats")
run_stats2 = outcar2.pop("run_stats")
self.assertEqual(len(doc["calcs_reversed"]), 2)
self.assertEqual(doc["composition_reduced"], {"Si": 1.0})
self.assertEqual(doc["composition_unit_cell"], {"Si": 2.0})
self.assertAlmostEqual(doc["output"]["energy"], oszicar2.ionic_steps[-1]["E0"])
self.assertEqual(doc["formula_pretty"], "Si")
self.assertEqual(doc["formula_anonymous"], "A")
self.assertEqual(list(doc["calcs_reversed"][0]["input"].keys()), list(doc["calcs_reversed"][1]["input"].keys()))
self.assertEqual(
list(doc["calcs_reversed"][0]["output"].keys()), list(doc["calcs_reversed"][1]["output"].keys())
)
self.assertEqual(doc["calcs_reversed"][0]["output"]["energy"], doc["output"]["energy"])
self.assertEqual(doc["run_stats"][doc["calcs_reversed"][0]["task"]["name"]], run_stats2)
self.assertEqual(doc["run_stats"][doc["calcs_reversed"][1]["task"]["name"]], run_stats1)
self.assertEqual(doc["calcs_reversed"][0]["output"]["outcar"], outcar2)
self.assertEqual(doc["calcs_reversed"][1]["output"]["outcar"], outcar1)
def test_runs_assimilate(self):
drone = VaspDrone(runs=["relax1", "relax2"])
doc = drone.assimilate(self.relax2)
oszicar2 = Oszicar(os.path.join(self.relax2, "OSZICAR.relax2.gz"))
outcar1 = Outcar(os.path.join(self.relax2, "OUTCAR.relax1.gz"))
outcar2 = Outcar(os.path.join(self.relax2, "OUTCAR.relax2.gz"))
outcar1 = outcar1.as_dict()
outcar2 = outcar2.as_dict()
run_stats1 = outcar1.pop("run_stats")
run_stats2 = outcar2.pop("run_stats")
self.assertEqual(len(doc["calcs_reversed"]), 2)
self.assertEqual(doc["composition_reduced"], {"Si": 1.0})
self.assertEqual(doc["composition_unit_cell"], {"Si": 2.0})
self.assertAlmostEqual(doc["output"]["energy"], oszicar2.ionic_steps[-1]["E0"])
self.assertEqual(doc["formula_pretty"], "Si")
self.assertEqual(doc["formula_anonymous"], "A")
self.assertEqual(
list(doc["calcs_reversed"][0]["input"].keys()),
list(doc["calcs_reversed"][1]["input"].keys()),
)
self.assertEqual(
list(doc["calcs_reversed"][0]["output"].keys()),
list(doc["calcs_reversed"][1]["output"].keys()),
)
self.assertEqual(
doc["calcs_reversed"][0]["output"]["energy"], doc["output"]["energy"]
)
self.assertEqual(
doc["run_stats"][doc["calcs_reversed"][0]["task"]["name"]], run_stats2
)
self.assertEqual(
doc["run_stats"][doc["calcs_reversed"][1]["task"]["name"]], run_stats1
)
self.assertEqual(doc["calcs_reversed"][0]["output"]["outcar"], outcar2)
self.assertEqual(doc["calcs_reversed"][1]["output"]["outcar"], outcar1)
def process_task(self, path):
try:
#Override incorrect outcar subdocs for two step relaxations
if os.path.exists(os.path.join(path, "relax2")):
try:
run_stats = {}
for i in [1,2]:
outcar = Outcar(zpath(os.path.join(path,"relax"+str(i), "OUTCAR")))
m_key = "calculations."+str(i-1)+".output.outcar"
self.tasks.update({'dir_name_full': path}, {'$set': {m_key: outcar.as_dict()}})
run_stats["relax"+str(i)] = outcar.run_stats
except:
logger.error("Bad OUTCAR for {}.".format(path))
try:
overall_run_stats = {}
for key in ["Total CPU time used (sec)", "User time (sec)",
"System time (sec)", "Elapsed time (sec)"]:
overall_run_stats[key] = sum([v[key]
for v in run_stats.values()])
run_stats["overall"] = overall_run_stats
except:
logger.error("Bad run stats for {}.".format(path))
self.tasks.update({'dir_name_full': path}, {'$set': {"run_stats": run_stats}})
print 'FINISHED', path
else:
print 'SKIPPING', path
except:
print '-----'
print 'ENCOUNTERED AN EXCEPTION!!!', path
traceback.print_exc()
print '-----'
def process_task(self, path):
try:
#Override incorrect outcar subdocs for two step relaxations
if os.path.exists(os.path.join(path, "relax2")):
try:
run_stats = {}
for i in [1, 2]:
outcar = Outcar(
zpath(
os.path.join(path, "relax" + str(i),
"OUTCAR")))
m_key = "calculations." + str(i - 1) + ".output.outcar"
self.tasks.update({'dir_name_full': path},
{'$set': {
m_key: outcar.as_dict()
}})
run_stats["relax" + str(i)] = outcar.run_stats
except:
logger.error("Bad OUTCAR for {}.".format(path))
try:
overall_run_stats = {}
for key in [
"Total CPU time used (sec)", "User time (sec)",
"System time (sec)", "Elapsed time (sec)"
]:
overall_run_stats[key] = sum(
[v[key] for v in run_stats.values()])
run_stats["overall"] = overall_run_stats
except:
logger.error("Bad run stats for {}.".format(path))
self.tasks.update({'dir_name_full': path},
{'$set': {
"run_stats": run_stats
}})
print 'FINISHED', path
else:
print 'SKIPPING', path
except:
print '-----'
print 'ENCOUNTERED AN EXCEPTION!!!', path
traceback.print_exc()
print '-----'
def post_process(self, dir_name, d):
"""
Simple post-processing for various files other than the vasprun.xml.
Called by generate_task_doc. Modify this if your runs have other
kinds of processing requirements.
Args:
dir_name:
The dir_name.
d:
Current doc generated.
"""
logger.info("Post-processing dir:{}".format(dir_name))
fullpath = os.path.abspath(dir_name)
# VASP input generated by pymatgen's alchemy has a
# transformations.json file that keeps track of the origin of a
# particular structure. This is extremely useful for tracing back a
# result. If such a file is found, it is inserted into the task doc
# as d["transformations"]
transformations = {}
filenames = glob.glob(os.path.join(fullpath, "transformations.json*"))
if len(filenames) >= 1:
with zopen(filenames[0], "rt") as f:
transformations = json.load(f)
try:
m = re.match("(\d+)-ICSD",
transformations["history"][0]["source"])
if m:
d["icsd_id"] = int(m.group(1))
except Exception as ex:
logger.warning("Cannot parse ICSD from transformations "
"file.")
pass
else:
logger.warning("Transformations file does not exist.")
other_parameters = transformations.get("other_parameters")
new_tags = None
if other_parameters:
# We don't want to leave tags or authors in the
# transformations file because they'd be copied into
# every structure generated after this one.
new_tags = other_parameters.pop("tags", None)
new_author = other_parameters.pop("author", None)
if new_author:
d["author"] = new_author
if not other_parameters: # if dict is now empty remove it
transformations.pop("other_parameters")
d["transformations"] = transformations
# Calculations done using custodian has a custodian.json,
# which tracks the jobs performed and any errors detected and fixed.
# This is useful for tracking what has actually be done to get a
# result. If such a file is found, it is inserted into the task doc
# as d["custodian"]
filenames = glob.glob(os.path.join(fullpath, "custodian.json*"))
if len(filenames) >= 1:
with zopen(filenames[0], "rt") as f:
d["custodian"] = json.load(f)
# Parse OUTCAR for additional information and run stats that are
# generally not in vasprun.xml.
try:
run_stats = {}
for filename in glob.glob(os.path.join(fullpath, "OUTCAR*")):
outcar = Outcar(filename)
i = 1 if re.search("relax2", filename) else 0
taskname = "relax2" if re.search("relax2", filename) else \
"relax1"
d["calculations"][i]["output"]["outcar"] = outcar.as_dict()
run_stats[taskname] = outcar.run_stats
except:
logger.error("Bad OUTCAR for {}.".format(fullpath))
try:
overall_run_stats = {}
for key in [
"Total CPU time used (sec)", "User time (sec)",
"System time (sec)", "Elapsed time (sec)"
]:
overall_run_stats[key] = sum(
[v[key] for v in run_stats.values()])
run_stats["overall"] = overall_run_stats
except:
logger.error("Bad run stats for {}.".format(fullpath))
d["run_stats"] = run_stats
#Convert to full uri path.
if self.use_full_uri:
d["dir_name"] = get_uri(dir_name)
if new_tags:
d["tags"] = new_tags
logger.info("Post-processed " + fullpath)
def post_process(self, dir_name, d):
"""
Simple post-processing for various files other than the vasprun.xml.
Called by generate_task_doc. Modify this if your runs have other
kinds of processing requirements.
Args:
dir_name:
The dir_name.
d:
Current doc generated.
"""
logger.info("Post-processing dir:{}".format(dir_name))
fullpath = os.path.abspath(dir_name)
# VASP input generated by pymatgen's alchemy has a
# transformations.json file that keeps track of the origin of a
# particular structure. This is extremely useful for tracing back a
# result. If such a file is found, it is inserted into the task doc
# as d["transformations"]
transformations = {}
filenames = glob.glob(os.path.join(fullpath, "transformations.json*"))
if len(filenames) >= 1:
with zopen(filenames[0], "rt") as f:
transformations = json.load(f)
try:
m = re.match("(\d+)-ICSD",
transformations["history"][0]["source"])
if m:
d["icsd_id"] = int(m.group(1))
except Exception as ex:
logger.warning("Cannot parse ICSD from transformations "
"file.")
pass
else:
logger.warning("Transformations file does not exist.")
other_parameters = transformations.get("other_parameters")
new_tags = None
if other_parameters:
# We don't want to leave tags or authors in the
# transformations file because they'd be copied into
# every structure generated after this one.
new_tags = other_parameters.pop("tags", None)
new_author = other_parameters.pop("author", None)
if new_author:
d["author"] = new_author
if not other_parameters: # if dict is now empty remove it
transformations.pop("other_parameters")
d["transformations"] = transformations
# Calculations done using custodian has a custodian.json,
# which tracks the jobs performed and any errors detected and fixed.
# This is useful for tracking what has actually be done to get a
# result. If such a file is found, it is inserted into the task doc
# as d["custodian"]
filenames = glob.glob(os.path.join(fullpath, "custodian.json*"))
if len(filenames) >= 1:
with zopen(filenames[0], "rt") as f:
d["custodian"] = json.load(f)
# Parse OUTCAR for additional information and run stats that are
# generally not in vasprun.xml.
try:
run_stats = {}
for filename in glob.glob(os.path.join(fullpath, "OUTCAR*")):
outcar = Outcar(filename)
i = 1 if re.search("relax2", filename) else 0
taskname = "relax2" if re.search("relax2", filename) else \
"relax1"
d["calculations"][i]["output"]["outcar"] = outcar.as_dict()
run_stats[taskname] = outcar.run_stats
except:
logger.error("Bad OUTCAR for {}.".format(fullpath))
try:
overall_run_stats = {}
for key in ["Total CPU time used (sec)", "User time (sec)",
"System time (sec)", "Elapsed time (sec)"]:
overall_run_stats[key] = sum([v[key]
for v in run_stats.values()])
run_stats["overall"] = overall_run_stats
except:
logger.error("Bad run stats for {}.".format(fullpath))
d["run_stats"] = run_stats
#Convert to full uri path.
if self.use_full_uri:
d["dir_name"] = get_uri(dir_name)
if new_tags:
d["tags"] = new_tags
logger.info("Post-processed " + fullpath)
def band_structure():
check_matplotlib()
step_count=1
filename='vasprun.xml'
check_file(filename)
proc_str="Reading Data From "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
vsr=Vasprun(filename)
step_count+=1
filename='KPOINTS'
check_file(filename)
proc_str="Reading Data From "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
bands = vsr.get_band_structure(filename, line_mode=True, efermi=vsr.efermi)
step_count+=1
filename='OUTCAR'
check_file(filename)
proc_str="Reading Data From "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
outcar=Outcar('OUTCAR')
mag=outcar.as_dict()['total_magnetization']
if vsr.is_spin:
proc_str="This Is a Spin-polarized Calculation."
procs(proc_str,0,sp='-->>')
tdos=vsr.tdos
SpinUp_gap=tdos.get_gap(spin=Spin.up)
cbm_vbm_up=tdos.get_cbm_vbm(spin=Spin.up)
SpinDown_gap=tdos.get_gap(spin=Spin.down)
cbm_vbm_down=tdos.get_cbm_vbm(spin=Spin.up)
if SpinUp_gap > min_gap and SpinDown_gap > min_gap:
is_metal=False
is_semimetal=False
elif SpinUp_gap > min_gap and SpinDown_gap < min_gap:
is_metal=False
is_semimetal=True
elif SpinUp_gap < min_gap and SpinDown_gap > min_gap:
is_metal=False
is_semimetal=True
elif SpinUp_gap < min_gap and SpinDown_gap < min_gap:
is_metal=True
is_semimetal=False
if is_metal:
proc_str="This Material Is a Metal."
procs(proc_str,0,sp='-->>')
if not is_metal and is_semimetal:
proc_str="This Material Is a Semimetal."
procs(proc_str,0,sp='-->>')
else:
proc_str="This Material Is a Semiconductor."
procs(proc_str,0,sp='-->>')
proc_str="Total magnetization is "+str(mag)
procs(proc_str,0,sp='-->>')
if mag > min_mag:
proc_str="SpinUp : vbm=%f eV cbm=%f eV gap=%f eV"%(cbm_vbm_up[1],cbm_vbm_up[0],SpinUp_gap)
procs(proc_str,0,sp='-->>')
proc_str="SpinDown: vbm=%f eV cbm=%f eV gap=%f eV"%(cbm_vbm_down[1],cbm_vbm_down[0],SpinUp_gap)
procs(proc_str,0,sp='-->>')
else:
proc_str="SpinUp : vbm=%f eV cbm=%f eV gap=%f eV"%(cbm_vbm_up[1],cbm_vbm_up[0],SpinUp_gap)
procs(proc_str,0,sp='-->>')
step_count+=1
filename="BAND.dat"
proc_str="Writting Band Structure Data to "+ filename +" File ..."
procs(proc_str,step_count,sp='-->>')
band_data_up=bands.bands[Spin.up]
band_data_down=bands.bands[Spin.down]
y_data_up=band_data_up.reshape(1,band_data_up.shape[0]*band_data_up.shape[1])[0]-vsr.efermi #shift fermi level to 0
y_data_down=band_data_down.reshape(1,band_data_down.shape[0]*band_data_down.shape[1])[0]-vsr.efermi #shift fermi level to 0
x_data=np.array(bands.distance*band_data_up.shape[0])
data=np.vstack((x_data,y_data_up,y_data_down)).T
head_line="#%(key1)+12s%(key2)+13s%(key3)+15s"%{'key1':'K-Distance','key2':'UpEnergy(ev)','key3':'DownEnergy(ev)'}
write_col_data(filename,data,head_line,band_data_up.shape[1])
else:
if vsr.parameters['LNONCOLLINEAR']:
proc_str="This Is a Non-Collinear Calculation."
else:
proc_str="This Is a Non-Spin Calculation."
procs(proc_str,0,sp='-->>')
cbm=bands.get_cbm()['energy']
vbm=bands.get_vbm()['energy']
gap=bands.get_band_gap()['energy']
if not bands.is_metal():
proc_str="This Material Is a Semiconductor."
procs(proc_str,0,sp='-->>')
proc_str="vbm=%f eV cbm=%f eV gap=%f eV"%(vbm,cbm,gap)
procs(proc_str,0,sp='-->>')
else:
proc_str="This Material Is a Metal."
procs(proc_str,0,sp='-->>')
step_count+=1
filename3="BAND.dat"
proc_str="Writting Band Structure Data to "+ filename3 +" File ..."
procs(proc_str,step_count,sp='-->>')
band_data=bands.bands[Spin.up]
y_data=band_data.reshape(1,band_data.shape[0]*band_data.shape[1])[0]-vsr.efermi #shift fermi level to 0
x_data=np.array(bands.distance*band_data.shape[0])
data=np.vstack((x_data,y_data)).T
head_line="#%(key1)+12s%(key2)+13s"%{'key1':'K-Distance','key2':'Energy(ev)'}
write_col_data(filename3,data,head_line,band_data.shape[1])
step_count+=1
bsp=BSPlotter(bands)
filename4="HighSymmetricPoints.dat"
proc_str="Writting Label infomation to "+ filename4 +" File ..."
procs(proc_str,step_count,sp='-->>')
head_line="#%(key1)+12s%(key2)+12s%(key3)+12s"%{'key1':'index','key2':'label','key3':'position'}
line=head_line+'\n'
for i,label in enumerate(bsp.get_ticks()['label']):
new_line="%(key1)12d%(key2)+12s%(key3)12f\n"%{'key1':i,'key2':label,'key3':bsp.get_ticks()['distance'][i]}
line+=new_line
line+='\n'
write_col_data(filename4,line,'',str_data=True)
try:
step_count+=1
filename5="BAND.png"
proc_str="Saving Plot to "+ filename5 +" File ..."
procs(proc_str,step_count,sp='-->>')
bsp.save_plot(filename5, img_format="png")
except:
print("Figure output fails !!!")
