Output Files

AZURE2 produces several output files in the configured output directory. An important distinction: while all input is in the laboratory frame, all quantities in output files are in the center-of-mass frame.

Parameter Files

param.par

Contains the initial formal R-matrix parameters (energies, reduced width amplitudes, etc.) derived from the Input File. Primarily a check file with limited direct use.

param.sav

Contains the best-fit formal R-matrix parameters after a fit is completed. This file can be loaded back into AZURE2 to reproduce a fit or to use as starting parameters for subsequent calculations (extrapolation, reaction rate, etc.).

parameters.out

Contains the physical (observable) parameters resulting from the fit. If the user wishes to use these as new starting values, they must be manually entered into the Levels and Channels tab.

normalizations.out

Contains the fitted normalization factors for data segments where normalization was varied. This file is automatically loaded when param.sav is selected.

Cross Section Output

AZUREOut_aa=_R=.out

Output from Calculate With Data and Fit With Data modes. The filename encodes the entrance (aa) and exit (R) particle pair indices.

Nine columns:

Col.	Description
1	Center-of-mass energy (MeV)
2	Excitation energy (MeV)
3	Center-of-mass angle (degrees)
4	Fit center-of-mass cross section (barns or barns/sr), or the fitted analyzing power (dimensionless) for an analyzing-power segment
5	Fit center-of-mass S-factor (MeV b or MeV b/sr)
6	Data center-of-mass cross section (barns or barns/sr), or the measured analyzing power (dimensionless) for an analyzing-power segment
7	Data center-of-mass cross section uncertainty (barns or barns/sr)
8	Data center-of-mass S-factor (MeV b or MeV b/sr)
9	Data center-of-mass S-factor uncertainty (MeV b or MeV b/sr)

When multiple segments share the same entrance and exit particle pairs, their data are written to the same file in the order they appear in the Segments tab, separated by a double blank line.

AZUREOut_aa=_R=.extrap

Output from Calculate Segments Without Data mode. Same naming convention as above. Five columns:

Col.	Description
1	Center-of-mass energy (MeV)
2	Excitation energy (MeV)
3	Center-of-mass angle (degrees)
4	Extrapolated center-of-mass cross section (barns or barns/sr), or the analyzing power (dimensionless) for an analyzing-power segment
5	Extrapolated center-of-mass S-factor (MeV b or MeV b/sr)

Uncertainty and Statistics

chiSquared.out

One line per data segment, then a total:

Segment#, Chi-Squared,  N,  Norm,  Norm-Chi-Squared
1,823.88,17,1,0
...
Total-Chi-Squared: 107456 Total-Norm-Chi-Squared: 0 Total-N: 415

Chi-Squared and Total-Chi-Squared are the data term only; Norm-Chi-Squared is the separate penalty on a varied normalization, and N counts data points (not degrees of freedom). The quantity a fit actually minimises is the sum of both — see How the Chi-Squared Is Built.

This file is the quickest scalar check that a run succeeded.

param.errors

Contains the reduced width amplitudes and their asymmetric uncertainties from a MINOS error analysis.

covariance_matrix.out

Contains the covariance and correlation matrices from a MINOS calculation, providing a complete description of parameter correlations.

Other Files

intEC.dat

External capture integral values for data segments. This file can be reused to speed up subsequent calculations, as long as:

No calculation segments have been added or removed.
No levels of a new \(J^\pi\) have been added or removed.
No channels have been added or removed.
The channel radius has not changed.

Level energies, widths, and ANCs can be changed freely while reusing this file.

intEC.extrap

Same as intEC.dat, but for the calculation (extrapolation) segments.

reactionrates.dat

Contains temperatures (in GK) and calculated reaction rates (in cm³ mol^-1 s^-1) from the Calculate Reaction Rate mode.

samples.mcmc

The MCMC chain, as CSV, one row per walker per step:

step,walker,logprob,loglikelihood,logprior,param0,param1,...

Every accepted state appears exactly once, including the repeated states a rejected proposal contributes — that repetition is how a Markov chain carries probability mass, so the file must not be deduplicated. logprob equals loglikelihood + logprior exactly. See MCMC Tab for how to load it and what to check before using it.

walkers.mcmc

The final position of every walker, written at the end of an MCMC run and when one is stopped early. Its purpose is resuming: with it, a continued run picks the ensemble up where it left off instead of re-scattering the walkers and splicing a fresh burn-in into the middle of the chain.