Log of /source

Add asserts that numbers are finite.

Add an option to force highest levels to LTE.  

Add some fflush statements. Redistribute n-manifold recomb above some n (100 for now) according to stat weight of recombined levels nLSJ.

cleanup unused variables.

some minor cleanup

Finally works.  Here are a few changes to print total Aul instead of average (which is not physically meaningful)

up_J and lo_J were ill-defined.  Use L+S

Nothing was printed because indices had not been filled.  Fix.

Simplify this loop dramatically.

Correct a few problems with printout in Helium.c and print_Levels.c. For the purposes of comparison with Cloudy results, LS multiplets were defined as {nLS_lo, nL_hi}.  This is incorrect because we want the rate *out* of nLS.  Changes in print_Results_Ryan.c reflect this.  Before this change 4^SF-3^SD showed 50% errors as compared with Cloudy.  Adding up ST results by hand yielded only small differences compared with Cloudy. 

more adjustments to print

comment out unused code

kill hack for D - don't understand why it's there, but problem emerges at n=26 because it requires hydrogenic at n=25 to have been calculated already.  With HS98 now covering n<=25, that no longer happens.  Just don't use the hack, and the scaling uses HS98 factors.  Also strengthen a few more asserts.

long int -> int in print format.

add a bunch of asserts

remove Ryan_Topbase_Fits - totally superseded by HS98

bring in HS98 cross-sections

These changes bring the code to "Fix 1" as stored on my web page at www.pa.uky.edu/~rporter/j-resolved/.  The following description was recorded:

Fix 1:
--Bug in Hummer_Storey_Recomb_Factor_25 caused the code to fail for some low temperatures, reported by Leticia Martin Hernandez.
--Added high and low temperature bounds.  Code will complain and exit if bounds are exceeded.

This is the original source provided to me by Rob Bauman.  -- Ryan Porter