def full_report(self):
N = self.adsorbophore_db.session.query(SQL_Adsorbophore).count()
del self._adsorbophores
del self._active_sites
csv = CSV('full_report.csv')
csv.set_headings("rank","N_sites", "N_unq_MOFs", "rmsd_error", "av_elstat", "stdev_elstat", "av_vdw", "stdev_vdw")
for rank in range(N):
adsorbophore = self.adsorbophore_from_sql(rank)
nsites = len(adsorbophore.active_sites)
mofs = {}
for site in adsorbophore.active_sites:
mofname = '.'.join(site.name.split('.')[:-1])
mofs.setdefault(mofname, 0)
mofs[mofname] += 1
error, vdw_mean, vdw_std, el_mean, el_std = self.obtain_error(rank)
csv.add_data(**{'rank.1': rank, 'N_sites.1':nsites, 'N_unq_MOFs.1':len(mofs.keys()),
'rmsd_error.1':error, 'av_elstat.1':el_mean, 'stdev_elstat.1':el_std,
'av_vdw.1':vdw_mean, 'stdev_vdw.1':vdw_std})
csv.write()
def _sbu_report(self):
"""Compute the surface areas and report them to a .csv file."""
# WARNING - this assumes that SBUs with the same name but in
# different topologies are the same, and will take the last instance
met_sbus = {}
org_sbus = {}
for sbu in self.sbu_pool:
if sbu.is_metal:
met_sbus[sbu.name] = sbu
else:
org_sbus[sbu.name] = sbu
filename = os.path.join(self.options.job_dir,
self.options.jobname + ".SBU_report.csv")
report = CSV(name=filename)
report.set_headings("sbu_id")
if self.options.calc_sbu_surface_area:
report.set_headings("surface_area")
if self.options.calc_max_sbu_span:
report.set_headings("sbu_span")
# metal sbus first.
for name, sbu in met_sbus.items():
info("Computing data for %s"%name)
report.add_data(**{"sbu_id.1": sbu.identifier})
if self.options.calc_sbu_surface_area:
report.add_data(**{"surface_area.1": sbu.surface_area})
if self.options.calc_max_sbu_span:
report.add_data(**{"sbu_span.1":sbu.max_span})
# list organic SBUs second.
for name, sbu in org_sbus.items():
info("Computing data for %s"%name)
report.add_data(**{"sbu_id.1": sbu.identifier})
if self.options.calc_sbu_surface_area:
report.add_data(**{"surface_area.1": sbu.surface_area})
if self.options.calc_max_sbu_span:
report.add_data(**{"sbu_span.1": sbu.max_span})
report.write()
def _sbu_report(self):
"""Compute the surface areas and report them to a .csv file."""
# WARNING - this assumes that SBUs with the same name but in
# different topologies are the same, and will take the last instance
met_sbus = {}
org_sbus = {}
for sbu in self.sbu_pool:
if sbu.is_metal:
met_sbus[sbu.name] = sbu
else:
org_sbus[sbu.name] = sbu
filename = os.path.join(self.options.job_dir,
self.options.jobname + ".SBU_report.csv")
report = CSV(name=filename)
report.set_headings("sbu_id")
if self.options.calc_sbu_surface_area:
report.set_headings("surface_area")
if self.options.calc_max_sbu_span:
report.set_headings("sbu_span")
# metal sbus first.
for name, sbu in met_sbus.items():
info("Computing data for %s" % name)
report.add_data(**{"sbu_id.1": sbu.identifier})
if self.options.calc_sbu_surface_area:
report.add_data(**{"surface_area.1": sbu.surface_area})
if self.options.calc_max_sbu_span:
report.add_data(**{"sbu_span.1": sbu.max_span})
# list organic SBUs second.
for name, sbu in org_sbus.items():
info("Computing data for %s" % name)
report.add_data(**{"sbu_id.1": sbu.identifier})
if self.options.calc_sbu_surface_area:
report.add_data(**{"surface_area.1": sbu.surface_area})
if self.options.calc_max_sbu_span:
report.add_data(**{"sbu_span.1": sbu.max_span})
report.write()
def obtain_rdfs(self, rank):
"""Return the radial distribution functions of the cliques in the original MOFs,
with the cliques cut-out."""
adsorbophore = self.adsorbophore_from_sql(rank)
nconfig = 0 # configurations counted
nparticles = {} # number of particles counted
distances = {} # keep distances in lists.
densities = {}
for site in adsorbophore.active_sites:
indices = [i.index for i in site.indices]
act_site = self.active_site_from_sql(site.name)
ads_atoms = [act_site.atoms[i] for i in indices]
site_eng = act_site.vdweng + act_site.eleng
if site_eng <= self.options.en_max and site_eng >= self.options.en_min:
mofpath = act_site.mofpath
# get mof from cif
mofname = os.path.split(mofpath)[-1][:-4]
mof = Structure(mofname)
mof.from_cif(mofpath)
# get atoms to cut out of mof
cut_inds = [atom.mof_id for atom in ads_atoms]
cut_coords = np.array([np.array((atom.x, atom.y, atom.z)) for atom in ads_atoms])
# get co2
co2 = self.return_co2_array(act_site)
rdf_centre = co2[0] # or self.centre_of_atoms(np.array([np.array([a.x, a.y, a.z]) for a in ads_atoms]))
# compute the minimum image convention of a supercell around the rdf centre.
# cut out the atoms in the active site SQL_ActiveSiteAtoms.mof_id
for atom in ads_atoms:
rdf_centre = np.array([atom.x, atom.y, atom.z])
original_indices, coordinates = self.min_img(mof, rdf_centre, (3,3,3), cut_coords)#cut_inds)
dists = distance.cdist(np.column_stack(rdf_centre), coordinates)
# debug
#f = open('debug.xyz', 'a')
#f.writelines("%i\n%s\n"%(len(coordinates)+1, "test"))
#for id, coord in enumerate(coordinates):
# orig_ind = original_indices[id]
# element = mof.atoms[orig_ind].type
# f.writelines("%s %9.5f %9.5f %9.5f\n"%(element, coord[0], coord[1], coord[2]))
#f.writelines("%s %9.5f %9.5f %9.5f\n"%("As", co2[0][0], co2[0][1], co2[0][2]))
#f.close()
nconfig += 1
for (x,y), val in np.ndenumerate(dists):
orig_ind = original_indices[y]
element = mof.atoms[orig_ind].uff_type
distances.setdefault(element, []).append(val)
nparticles.setdefault(element, 0)
nparticles[element] += 1
# add number densities from unit cell.
counts = {}
for atom in mof.atoms:
counts.setdefault(atom.uff_type, 0)
counts[atom.uff_type] += 1
for element, count in counts.items():
densities.setdefault(element, []).append(count/mof.cell.volume)
# compute RDFs by atom type? element? general?
for element, dis in distances.items():
rho = np.mean(densities[element])
hist, bin_edges = np.histogram(dis, bins=self.options.rdf_bins,
range=(0., self.options.rdf_dist))
dr = np.diff(bin_edges)[0]
norm = rho * float(nconfig)
#shell_volume = 4/3*pi*pow(r, 2)*dr
#rho, norm = 1., 1.
rdf = [hist[i]/norm/(4./3.*pi*(pow((i+0.5)*dr, 3) - pow((i-0.5)*dr, 3)))
for i in range(self.options.rdf_bins)]
Emax = self.options.en_max
Emin = self.options.en_min
el_name = '.'.join([j for j in element.split('_') if j])
csv = CSV("RDF_rank%i_%s_Emax_%0.2f_Emin_%0.2f"%(rank, el_name, Emax, Emin))
csv.set_headings("r", "g(r)")
for i, val in enumerate(rdf):
csv.add_data(**{"r.1":i*dr, "g(r).1":val})
csv.write()
