import iminuit
import logging
import numpy as np
log = logging.getLogger(__name__)
[docs]class minuit_optimizer(object):
[docs] def __init__(self, verbose=False, ncall=10000, errordef=1, steps=1000):
self.verbose = 0
self.ncall = ncall
self.errordef = errordef
self.steps = steps
def _make_minuit(
self, objective, data, pdf, init_pars, init_bounds, fixed_vals=None
):
def f(pars):
result = objective(pars, data, pdf)
logpdf = result[0]
return logpdf
parnames = ['p{}'.format(i) for i in range(len(init_pars))]
kw = {'limit_p{}'.format(i): b for i, b in enumerate(init_bounds)}
initvals = {'p{}'.format(i): v for i, v in enumerate(init_pars)}
step_sizes = {
'error_p{}'.format(i): (b[1] - b[0]) / float(self.steps)
for i, b in enumerate(init_bounds)
}
fixed_vals = fixed_vals or []
constraints = {}
for index, value in fixed_vals:
constraints = {'fix_p{}'.format(index): True}
initvals['p{}'.format(index)] = value
kwargs = {}
for d in [kw, constraints, initvals, step_sizes]:
kwargs.update(**d)
mm = iminuit.Minuit(
f,
print_level=1 if self.verbose else 0,
errordef=1,
use_array_call=True,
forced_parameters=parnames,
**kwargs
)
return mm
[docs] def minimize(self, objective, data, pdf, init_pars, par_bounds, fixed_vals=None):
mm = self._make_minuit(objective, data, pdf, init_pars, par_bounds, fixed_vals)
result = mm.migrad(ncall=self.ncall)
assert result
return np.asarray([x[1] for x in mm.values.items()])
[docs] def unconstrained_bestfit(self, objective, data, pdf, init_pars, par_bounds):
return self.minimize(objective, data, pdf, init_pars, par_bounds)
[docs] def constrained_bestfit(
self, objective, constrained_mu, data, pdf, init_pars, par_bounds
):
return self.minimize(
objective,
data,
pdf,
init_pars,
par_bounds,
[(pdf.config.poi_index, constrained_mu)],
)