def fit(dm, method='map', keys=gbd_keys(), iter=50000, burn=25000, thin=1, verbose=1,
dbname='model.pickle'):
""" Generate an estimate of the generic disease model parameters
using maximum a posteriori liklihood (MAP) or Markov-chain Monte
Carlo (MCMC)
Parameters
----------
dm : dismod3.DiseaseModel
the object containing all the data, priors, and additional
information (like input and output age-mesh)
method : string, optional
the parameter estimation method, either 'map' or 'mcmc'
keys : list, optional
a list of gbd keys for the parameters to fit; it can speed up
computation to holding some parameters constant while allowing
others to vary
iter : int, optional
burn : int, optional
thin : int, optional
parameters for the MCMC, which control how long it takes, and
how accurate it is
"""
if not hasattr(dm, 'vars'):
print 'initializing model vars... ',
dm.calc_effective_sample_size(dm.data)
dm.vars = setup(dm, keys)
print 'finished'
if method == 'map':
print 'initializing MAP object... ',
map_method = 'fmin_powell'
#map_method = 'fmin_l_bfgs_b'
mc.MAP([dm.vars[k] for k in keys if k.find('incidence') != -1]).fit(method=map_method, iterlim=500, tol=.01, verbose=verbose)
mc.MAP([dm.vars[k] for k in keys if k.find('remission') != -1]).fit(method=map_method, iterlim=500, tol=.01, verbose=verbose)
mc.MAP([dm.vars[k] for k in keys if
k.find('excess-mortality') != -1 or
k.find('m') != -1 or
k.find('mortality') != -1 or
k.find('relative-risk') != -1 or
k.find('bins') != -1]).fit(method=map_method, iterlim=500, tol=.01, verbose=verbose)
mc.MAP([dm.vars[k] for k in keys if
k.find('incidence') != -1 or
k.find('bins') != -1 or
k.find('prevalence') != -1]).fit(method=map_method, iterlim=500, tol=.01, verbose=verbose)
mc.MAP([dm.vars[k] for k in keys if
k.find('excess-mortality') != -1 or
k.find('m') != -1 or
k.find('mortality') != -1 or
k.find('relative-risk') != -1 or
k.find('bins') != -1 or
k.find('prevalence') != -1]).fit(method=map_method, iterlim=500, tol=.01, verbose=verbose)
dm.map = mc.MAP(dm.vars)
print 'finished'
try:
dm.map.fit(method=map_method, iterlim=500, tol=.001, verbose=verbose)
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
for k in keys:
try:
val = dm.vars[k]['rate_stoch'].value
dm.set_map(k, val)
except KeyError:
pass
if method == 'norm_approx':
dm.na = mc.NormApprox(dm.vars, eps=.0001)
try:
dm.na.fit(method='fmin_powell', iterlim=500, tol=.00001, verbose=verbose)
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
for k in keys:
if dm.vars[k].has_key('rate_stoch'):
dm.set_map(k, dm.vars[k]['rate_stoch'].value)
try:
dm.na.sample(1000, verbose=verbose)
for k in keys:
# TODO: rename 'rate_stoch' to something more appropriate
if dm.vars[k].has_key('rate_stoch'):
rate_model.store_mcmc_fit(dm, k, dm.vars[k])
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
elif method == 'mcmc':
# make pymc warnings go to stdout
import sys
mc.warnings.warn = sys.stdout.write
dm.mcmc = mc.MCMC(dm.vars, db='pickle', dbname=dbname)
for k in keys:
if 'dispersion_step_sd' in dm.vars[k]:
dm.mcmc.use_step_method(mc.Metropolis, dm.vars[k]['log_dispersion'],
proposal_sd=dm.vars[k]['dispersion_step_sd'])
if 'age_coeffs_mesh_step_cov' in dm.vars[k]:
dm.mcmc.use_step_method(mc.AdaptiveMetropolis, dm.vars[k]['age_coeffs_mesh'],
cov=dm.vars[k]['age_coeffs_mesh_step_cov'], verbose=0)
try:
dm.mcmc.sample(iter=iter, thin=thin, burn=burn, verbose=verbose)
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
dm.mcmc.db.commit()
for k in keys:
t,r,y,s = type_region_year_sex_from_key(k)
if t in ['incidence', 'prevalence', 'remission', 'excess-mortality', 'mortality', 'prevalence_x_excess-mortality']:
import neg_binom_model
neg_binom_model.store_mcmc_fit(dm, k, dm.vars[k])
elif t in ['relative-risk', 'duration', 'incidence_x_duration']:
import normal_model
normal_model.store_mcmc_fit(dm, k, dm.vars[k])
def fit(dm, method='map'):
""" Generate an estimate of the generic disease model parameters
using maximum a posteriori liklihood (MAP) or Markov-chain Monte
Carlo (MCMC)
Parameters
----------
dm : dismod3.DiseaseModel
the object containing all the data, priors, and additional
information (like input and output age-mesh)
method : string, optional
the parameter estimation method, either 'map' or 'mcmc'
Example
-------
>>> import dismod3
>>> import dismod3.generic_disease_model as model
>>> dm = dismod3.get_disease_model(1)
>>> model.fit(dm, method='map')
>>> model.fit(dm, method='mcmc')
"""
if not hasattr(dm, 'vars'):
for param_type in ['incidence', 'remission', 'excess-mortality']:
# find initial values for these rates
data = [d for d in dm.data if clean(d['data_type']).find(param_type) != -1]
# use a random subset of the data if there is a lot of it,
# to speed things up
if len(data) > 25:
dm.fit_initial_estimate(param_type, random.sample(data,25))
else:
dm.fit_initial_estimate(param_type, data)
dm.set_units(param_type, '(per person-year)')
dm.set_units('prevalence', '(per person)')
dm.set_units('duration', '(years)')
dm.vars = setup(dm)
if method == 'map':
if not hasattr(dm, 'map'):
dm.map = mc.MAP(dm.vars)
try:
dm.map.fit(method='fmin_powell', iterlim=500, tol=.001, verbose=1)
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
for t in dismod3.settings.output_data_types:
t = clean(t)
val = dm.vars[t]['rate_stoch'].value
dm.set_map(t, val)
dm.set_initial_value(t, val) # better initial value may save time in the future
elif method == 'mcmc':
if not hasattr(dm, 'mcmc'):
dm.mcmc = mc.MCMC(dm.vars)
for key in dm.vars:
stochs = dm.vars[key].get('logit_p_stochs', [])
if len(stochs) > 0:
dm.mcmc.use_step_method(mc.AdaptiveMetropolis, stochs)
try:
dm.mcmc.sample(iter=60*1000, burn=10*1000, thin=50, verbose=1)
except KeyboardInterrupt:
# if user cancels with cntl-c, save current values for "warm-start"
pass
for t in dismod3.settings.output_data_types:
t = clean(t)
rate_model.store_mcmc_fit(dm, t, dm.vars[t])
