def handle(self, *id_list, **options):
asrfs = AgeSpecificRateFunction.objects.filter(id__in=id_list)
#for asrf in asrfs:
# print "\n\nFast Fitting %s" % asrf.pk
# fit_rate_function.mcmc_fit(asrf, speed='fast')
for asrf in asrfs:
print "\n\nFitting %s" % asrf.id
fit_rate_function.mcmc_fit(asrf)
def handle(self, *args, **options):
if args[0] == "dm":
r = create_test_asrf(".05", rate_type="remission data", priors="smooth 10.0\nconfidence 1000 .1")
p = create_test_asrf(
".2+.3*.01*age", rate_type="prevalence data", priors="smooth 10.0\nincreasing 0 100\nconfidence 1000 .1"
)
cf = create_test_asrf("0.", rate_type="case fatality data", priors="smooth 10.0\nconfidence 200 .1")
dm = DiseaseModel(disease=cf.disease, region=cf.region, sex=cf.sex)
dm.save()
dm.rates = [r, p, cf]
dm.save()
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (dm.id, dm.get_absolute_url())
fit_disease_model.mcmc_fit(dm)
i = create_test_asrf("0.", rate_type="incidence data", priors="smooth 10.0\nconfidence 1000 .1")
p = create_test_asrf(
".5 - .3*.01*age",
rate_type="prevalence data",
priors="smooth 10.0\ndecreasing 0 100\nconfidence 1000 .1",
)
cf = create_test_asrf("0.", rate_type="case fatality data", priors="smooth 10.0\nconfidence 1000 .1")
dm = DiseaseModel(disease=cf.disease, region=cf.region, sex=cf.sex)
dm.save()
dm.rates = [i, p, cf]
dm.save()
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (dm.id, dm.get_absolute_url())
fit_disease_model.mcmc_fit(dm)
i = create_test_asrf("0.", rate_type="incidence data", priors="smooth 10.0\nconfidence 1000 .1")
r = create_test_asrf(".05", rate_type="remission data", priors="smooth 10.0\nconfidence 1000 .1")
cf = create_test_asrf("0.", rate_type="case fatality data", priors="smooth 10.0\nconfidence 1000 .1")
dm = DiseaseModel(disease=cf.disease, region=cf.region, sex=cf.sex)
dm.save()
dm.rates = [i, r, cf]
dm.save()
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (dm.id, dm.get_absolute_url())
fit_disease_model.mcmc_fit(dm)
i = create_test_asrf(".1", rate_type="incidence data", priors="smooth 10.0\nconfidence 1000 .1")
r = create_test_asrf(".05", rate_type="remission data", priors="smooth 10.0\nconfidence 1000 .1")
p = create_test_asrf(
".07*(age/100.0)**2", rate_type="prevalence data", priors="smooth 10.0\nconfidence 1000 .1"
)
dm = DiseaseModel(disease=cf.disease, region=cf.region, sex=cf.sex)
dm.save()
dm.rates = [i, r, p]
dm.save()
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (dm.id, dm.get_absolute_url())
fit_disease_model.mcmc_fit(dm)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (dm.id, dm.get_absolute_url())
elif args[0] == "rf":
if args[1] == "unimodal":
rf = create_test_asrf(
".6-2.*(.5-age/100.0)**2",
rate_type="incidence data",
priors="smooth 10.0\nunimodal 0 100\nconfidence 1000 .1",
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
rf = create_test_asrf(
".06-.2*(.5-age/100.0)**2", rate_type="incidence data", priors="smooth 10.0\nunimodal 0 100"
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
elif args[1] == "increasing":
rf = create_test_asrf(
".01 + .5*(age/100.0)**2",
rate_type="remission data",
priors="smooth 10.0\nincreasing 0 100\nconfidence 1000 .1",
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
rf = create_test_asrf(
".001 + .05*(age/100.0)", rate_type="incidence data", priors="smooth 10.0\nincreasing 0 100"
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
elif args[1] == "decreasing":
rf = create_test_asrf(
".61 - .5*(age/100.0)**2", rate_type="remission data", priors="smooth 10.0\ndecreasing 0 100"
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
rf = create_test_asrf(
".061 - .05*(age/100.0)", rate_type="incidence data", priors="smooth 10.0\ndecreasing 0 100"
)
print "Fitting %s\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
fit_rate_function.mcmc_fit(rf)
print "\nModel %d is fit\n http://winthrop.gs.washington.edu:5432%s" % (rf.id, rf.get_absolute_url())
def test_fit(self):
from dismod3.bayesian_models import fit_rate_function
from dismod3.tests import bayesian_probability_test
bayesian_probability_test.add_priors(self.asrf)
fit_rate_function.mcmc_fit(self.asrf, speed='testing fast')