TransientResult¶
- class TRXASprefitpack.driver.TransientResult[source]¶
Bases:
dictRepresent results for fitting driver routine
- model¶
model used for fitting
decay: sum of the convolution of exponential decay and instrumental response function
dmp_osc: sum of the convolution of damped oscillation and instrumental response function
both: sum of decay and dmp_osc model
- Type
{‘decay’, ‘dmp_osc’, ‘both’}
- name_of_dset¶
name of each dataset
- Type
sequence of str
- t¶
time range for each dataset
- Type
sequence of np.ndarray
- intensity¶
sequence of datasets of intensity of time delay scan
- Type
sequence of np.ndarray
- eps¶
sequence of datasets for estimated error of time delay scan
- Type
sequence of np.ndarray
- fit¶
fitting curve for each data set
- Type
sequence of np.ndarray
- fit_decay¶
decay part of fitting curve for each data set [model = ‘both’]
- Type
sequence of np.ndarray
- fit_osc¶
oscillation part of fitting curve for each data set [model = ‘both’]
- Type
sequence of np.ndarray
- res¶
residual curve (data-fit) for each data set
- Type
sequence of np.ndarray
- irf¶
shape of instrument response function
‘g’: gaussian instrumental response function
‘c’: cauchy (lorenzian) instrumental response function
‘pv’: pseudo voigt instrumental response function (linear combination of gaussian and lorenzian function)
- Type
{‘g’, ‘c’, ‘pv’}
- fwhm¶
unifrom fwhm parameter for pseudo voigt function \(((1-\eta)*g(t, {fwhm})+\eta*c{t, {fwhm}})\)
- Type
float
- eta¶
mixing parameter for pseudo voigt function \(((1-\eta)*g(t, {fwhm})+\eta*c{t, {fwhm}})\)
- Type
float
- param_name¶
name of parameter
- Type
np.ndarray
- n_decay¶
number of decay components (except baseline feature)
- Type
int
- n_osc¶
number of damped oscillation components
- Type
int
- x¶
best parameter
- Type
np.ndarray
- bounds¶
boundary of each parameter
- Type
sequence of tuple
- base¶
whether or not use baseline feature in fitting process
- Type
bool
- c¶
best weight of each component of each datasets
- Type
sequence of np.ndarray
- phase¶
phase factor of each oscillation component of each datasets [mode = ‘dmp_osc’, ‘both’]
- Type
sequence of np.ndarray
- chi2¶
total chi squared value of fitting
- Type
float
- aic¶
Akaike Information Criterion statistic: \(N\log(\chi^2/N)+2N_{parm}\)
- Type
float
- bic¶
Bayesian Information Criterion statistic: \(N\log(\chi^2/N)+N_{parm}\log(N)\)
- Type
float
- chi2_ind¶
chi squared value of individual time delay scan
- Type
np.ndarray
- red_chi2¶
total reduced chi squared value of fitting
- Type
float
- red_chi2_ind¶
reduced chi squared value of individul time delay scan
- Type
np.ndarray
- nfev¶
total number of function evaluation
- Type
int
- n_param¶
total number of effective parameter
- Type
int
- n_param_ind¶
number of parameter which affects fitting quality of indiviual time delay scan
- Type
int
- num_pts¶
total number of data points
- Type
int
- jac¶
jacobian of objective function at optimal point
- Type
np.ndarray
- cov¶
covariance matrix (i.e. inverse of (jac.T @ jac))
- Type
np.ndarray
- cov_scaled¶
scaled covariance matrix (i.e. red_chi2 * cov)
- Type
np.ndarray
- corr¶
parameter correlation matrix
- Type
np.ndarray
- x_eps¶
estimated error of parameter (i.e. square root of diagonal element of conv_scaled)
- Type
np.ndarray
- method_glb¶
method of global optimization used in fitting process
- Type
{‘basinhopping’}
- message_glb¶
messages from global optimization process
- Type
str
- method_lsq¶
method of local optimization for least_squares minimization (refinement of global optimization solution)
- Type
{‘trf’, ‘dogbox’, ‘lm’}
- success_lsq¶
whether or not local least square optimization is successed
- Type
bool
- message_lsq¶
messages from local least square optimization process
- Type
str
- status¶
status of optimization process
0 : normal termination
-1 : least square optimization process is failed
- Type
{0, -1}