def get_parameter_info(id=None):
"""Returns the parameter information needed for calling mht.
This routine will call covariance() if needed, but not
fit().
For now only works with a single dataset, and requires
the Cash statistic.
"""
if id == None:
idval = ui.get_default_id()
else:
idval = id
# Employ a lot of safety checks
#
fr = ui.get_fit_results()
if fr == None:
raise RuntimeError, "No fit results available!"
if len(fr.datasets) != 1:
raise RuntimeError, "Fit is for multiple datasets (%s) which we do not support!" % fr.datasets
if fr.datasets[0] != idval:
raise RuntimeError, "Fit results are for dataset %s, not %s" % (fr.datasets[0], idval)
if fr.statname != "cash":
raise RuntimeError, "Fit was run using statistic=%s rather than cash!" % fr.statname
if not fr.succeeded:
# Should use standard sherpa logging
print "Warning: fit to dataset %s did not complete successfully:\n%s" % (idval, fr.message)
cr = ui.get_covar_results()
if cr == None or len(cr.datasets) != 1 or cr.datasets[0] != idval:
# Should use standard sherpa logging
print "Running covariance for dataset %s" % idval
ui.covariance(idval)
cr = ui.get_covar_results()
if cr.statname != "cash":
raise RuntimeError, "Covariance was run using statistic=%s rather than cash!" % cr.statname
if len(fr.parnames) != len(cr.parnames):
raise RuntimeError, "Number of parameters used in fit and covariance analysis do not agree!\n fit=%s\n covar=%s\n" % (
fr.parnames,
cr.parnames,
)
for (p1, p2) in zip(fr.parnames, cr.parnames):
if p1 != p2:
raise RuntimeError, "Order of fit/covariance parameters does not match: %s vs %s" % (p1, p2)
for (pname, v1, v2) in zip(fr.parnames, fr.parvals, cr.parvals):
if v1 != v2:
raise RuntimeError, "Value of fit/covariance parameters does not match for parameter %s: %g vs %g" % (
pname,
v1,
v2,
)
if not hasattr(cr, "extra_output") or cr.extra_output == None:
raise RuntimeError, "get_covar_results has no .extra_output or it is None; is this CIAOX?"
# Store the information, we explicitly copy all items to avoid
# problems if fit/covariance are run again. This is done by converting
# all tuples to numpy arrays, even for strings, and is actually
# not needed.
#
out = {
"dataset": idval,
"npars": len(fr.parnames),
"parnames": np.asarray(fr.parnames),
"parvals": np.asarray(fr.parvals),
"parmins": np.asarray(cr.parmins),
"parmaxes": np.asarray(cr.parmaxes),
"sigma": cr.sigma,
"covar": cr.extra_output.copy(),
"statval": fr.statval,
}
return out
b1, b2 = pars
return BoxBOD.model(x, b1, b2)
ui.load_user_model(boxbod_func, "boxbod")
ui.add_user_pars("boxbod", ["b1", "b2"])
bb = boxbod
ui.set_model(bb)
bb.b1, bb.b2 = BoxBOD.p0[0], BoxBOD.p0[1]
# Perform fit
ui.set_stat('chi2datavar')
#ui.set_method('levmar') # ['levmar', 'moncar', 'neldermead', 'simplex']
ui.set_method_opt('xtol', 1e-10)
ui.fit() # Compute best-fit parameters
ui.set_covar_opt('eps', 1e-5) # @todo: Why does this parameter have no effect
ui.covariance() # Compute covariance matrix (i.e. errors)
#ui.conf() # Compute profile errors
#ui.show_all() # Print a very nice summary of your session to less
# Report results
fr = ui.get_fit_results()
cr = ui.get_covar_results()
# Report results (we have to apply the s factor ourselves)
popt = np.array(fr.parvals)
chi2 = fr.statval
s_factor = np.sqrt(chi2 / fr.dof)
perr = s_factor * np.array(cr.parmaxes)
report_results('sherpa', popt, perr, chi2)
def get_parameter_info(id=None):
"""Returns the parameter information needed for calling mht.
This routine will call covariance() if needed, but not
fit().
For now only works with a single dataset, and requires
the Cash statistic.
"""
if id == None:
idval = ui.get_default_id()
else:
idval = id
# Employ a lot of safety checks
#
fr = ui.get_fit_results()
if fr == None:
raise RuntimeError, "No fit results available!"
if len(fr.datasets) != 1:
raise RuntimeError, "Fit is for multiple datasets (%s) which we do not support!" % fr.datasets
if fr.datasets[0] != idval:
raise RuntimeError, "Fit results are for dataset %s, not %s" % (
fr.datasets[0], idval)
if fr.statname != "cash":
raise RuntimeError, "Fit was run using statistic=%s rather than cash!" % fr.statname
if not fr.succeeded:
# Should use standard sherpa logging
print "Warning: fit to dataset %s did not complete successfully:\n%s" % (
idval, fr.message)
cr = ui.get_covar_results()
if cr == None or len(cr.datasets) != 1 or cr.datasets[0] != idval:
# Should use standard sherpa logging
print "Running covariance for dataset %s" % idval
ui.covariance(idval)
cr = ui.get_covar_results()
if cr.statname != "cash":
raise RuntimeError, "Covariance was run using statistic=%s rather than cash!" % cr.statname
if len(fr.parnames) != len(cr.parnames):
raise RuntimeError, "Number of parameters used in fit and covariance analysis do not agree!\n fit=%s\n covar=%s\n" % (
fr.parnames, cr.parnames)
for (p1, p2) in zip(fr.parnames, cr.parnames):
if p1 != p2:
raise RuntimeError, "Order of fit/covariance parameters does not match: %s vs %s" % (
p1, p2)
for (pname, v1, v2) in zip(fr.parnames, fr.parvals, cr.parvals):
if v1 != v2:
raise RuntimeError, "Value of fit/covariance parameters does not match for parameter %s: %g vs %g" % (
pname, v1, v2)
if not hasattr(cr, "extra_output") or cr.extra_output == None:
raise RuntimeError, "get_covar_results has no .extra_output or it is None; is this CIAOX?"
# Store the information, we explicitly copy all items to avoid
# problems if fit/covariance are run again. This is done by converting
# all tuples to numpy arrays, even for strings, and is actually
# not needed.
#
out = {
"dataset": idval,
"npars": len(fr.parnames),
"parnames": np.asarray(fr.parnames),
"parvals": np.asarray(fr.parvals),
"parmins": np.asarray(cr.parmins),
"parmaxes": np.asarray(cr.parmaxes),
"sigma": cr.sigma,
"covar": cr.extra_output.copy(),
"statval": fr.statval
}
return out