lattice/pyerrors
1
0
Fork 0
mirror of https://github.com/fjosw/pyerrors.git synced 2025-05-15 20:13:41 +02:00
Code Issues Projects Releases Packages Wiki Activity

fix: index of num diff jacobian in least squares fit corrected.

Browse source
This commit is contained in:
Fabian Joswig 2022-10-05 17:54:25 +01:00
parent ee4149e498
commit 99e130d33c
No known key found for this signature in database
1 changed files with 10 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

11
pyerrors/fits.py
View file

@ -416,7 +416,10 @@ def _prior_fit(x, y, func, priors, silent=False, **kwargs):
if not m.fmin.is_valid: if not m.fmin.is_valid:
raise Exception('The minimization procedure did not converge.') raise Exception('The minimization procedure did not converge.')
hess_inv = np.linalg.pinv(jacobian(jacobian(chisqfunc))(params)) hess = jacobian(jacobian(chisqfunc))(params)
if kwargs.get('num_grad') is True:
hess = hess[0]
hess_inv = np.linalg.pinv(hess)
def chisqfunc_compact(d): def chisqfunc_compact(d):
model = func(d[:n_parms], x) model = func(d[:n_parms], x)
@ -424,6 +427,8 @@ def _prior_fit(x, y, func, priors, silent=False, **kwargs):
return chisq return chisq
jac_jac = jacobian(jacobian(chisqfunc_compact))(np.concatenate((params, y_f, p_f))) jac_jac = jacobian(jacobian(chisqfunc_compact))(np.concatenate((params, y_f, p_f)))
if kwargs.get('num_grad') is True:
jac_jac = jac_jac[0]
deriv = -hess_inv @ jac_jac[:n_parms, n_parms:] deriv = -hess_inv @ jac_jac[:n_parms, n_parms:]
@ -591,6 +596,8 @@ def _standard_fit(x, y, func, silent=False, **kwargs):
hess = jacobian(jacobian(chisqfunc))(fitp) hess = jacobian(jacobian(chisqfunc))(fitp)
except TypeError: except TypeError:
raise Exception("It is required to use autograd.numpy instead of numpy within fit functions, see the documentation for details.") from None raise Exception("It is required to use autograd.numpy instead of numpy within fit functions, see the documentation for details.") from None
if kwargs.get('num_grad') is True:
hess = hess[0]
if kwargs.get('correlated_fit') is True: if kwargs.get('correlated_fit') is True:
def chisqfunc_compact(d): def chisqfunc_compact(d):
@ -605,6 +612,8 @@ def _standard_fit(x, y, func, silent=False, **kwargs):
return chisq return chisq
jac_jac = jacobian(jacobian(chisqfunc_compact))(np.concatenate((fitp, y_f))) jac_jac = jacobian(jacobian(chisqfunc_compact))(np.concatenate((fitp, y_f)))
if kwargs.get('num_grad') is True:
jac_jac = jac_jac[0]
# Compute hess^{-1} @ jac_jac[:n_parms, n_parms:] using LAPACK dgesv # Compute hess^{-1} @ jac_jac[:n_parms, n_parms:] using LAPACK dgesv
try: try: