[esp-r] Re: [Core_developers] opinions requested about backward compatibilityin ESP-r results libraries

[Jon Hand]

Phylroy said...

::I am interested in what post-processing you are talking about. Does res 'massage' data? Or is it just the fact that the data ::reading from sql -> fortran 77 is problematic? Maybe we can discuss off-line? 

For an example of the post-processing that was being done look in esrures/castyp.F
which was needed because only the total sensible convective and total sensible
radiant casual gains was held for the zone.  To separate out the specifics for the
various reports hundreds of lines of code was used.   

There are some other places where some data is re-constituted (probably
related to surface convection).  The save level 2 does not have sufficient
information to present a zone energy balance so res mimics the calculations
done in the simulator.  We might decide this is a silly thing to do.

Another example of post-processing is reporting on MRT.  This requires
the surface temperatures in the room and so lots of reading is done
to gather the necessary data.  This is a different type of post-processing
to derive a new metric from a collection of existing data.

For this 2nd type of post-processing we might decide to let the simulator pre-calculate this
value and just write it - minor slow down in simulator and a faster results recovery.

In each of these cases, the initial desire to make a smaller results
file required that more work be done to expand the limited data.
In the future we could look at saving more performance criteria
so that the res module can become even more simple.

-Jon