From 1768699cad80e98f9f01806a1d8104d2b2aa9f7b Mon Sep 17 00:00:00 2001 From: Paul Romano Date: Mon, 15 Oct 2012 11:44:17 -0400 Subject: [PATCH] Added comment block describing state_point module. --- src/state_point.F90 | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/src/state_point.F90 b/src/state_point.F90 index e99358231c..f42599e355 100644 --- a/src/state_point.F90 +++ b/src/state_point.F90 @@ -1,5 +1,23 @@ module state_point +!=============================================================================== +! STATE_POINT -- This module handles writing and reading binary state point +! files. State points are contain complete tally results, source sites, and +! various other data. They can be used to restart a run or to reconstruct +! confidence intervals for tallies (this requires post-processing via Python +! scripts). +! +! Modifications to this module should be made with care. There are essentially +! three different ways to write or read state points: 1) normal Fortran file +! I/O, 2) MPI-IO, and 3) HDF5. The HDF5 functionality is contained in the +! hdf5_interface module. If you plan to change the state point, you will need to +! change all methods. You should also increment REVISION_STATEPOINT in the +! constants module. +! +! State points can be written at any batch during a simulation, or at specified +! intervals, using the tag. +!=============================================================================== + use error, only: warning, fatal_error use global use math, only: t_percentile