Log of /tags/release/c07.02.02/tsuite

tsuite/auto/runall.pl:
  Set nSkip to 0 so that entire test suite will be run.

merge rev 2002

first set of updates on slow test suite - all feii* sims corrected to values asserted by source at head of this branch.

H2* sims have not been run yet but changes are not expected

fix typo in asserted value

source/parse_table.cpp:
source/cdgetlinelist.cpp:
source/parse.h:
source/zero.cpp:
source/cont_createmesh.cpp:
source/parse_punch.cpp:
source/rfield.h:
source/optimize_phymir.cpp:
source/ion_recomb_Badnell.cpp:
source/cont_ffun.cpp:
source/cloudy.cpp:
source/cdinit.cpp:
source/cddrive.cpp:
source/atom_feii.cpp:
source/hydrobranch.cpp:
source/maincl.cpp:
  - Bug-fix: fix PR45 - large grid fails on max no. of open file descriptors exceeded.
    This is largely a port of r1623, r1711, r1726, and r1728. There are still remaining
    issues with "special" punch commands; these require a more comprehensive bugfix.
tsuite/auto/blr_rnfb.in:
tsuite/auto/blr_kk81.in:
tsuite/auto/blr_rnfa.in:
  - Fix near-botches, largely a result of r1471.

Merge various bug-fixes from the mainline:

r1439: fix convergence problems in stars_optimize[1-3].in

r1447: fix crash with g++ 4.2.x compilers.

r1454: fix compilation warnings on big-endian systems.

r1458: fix crashes with Sun Studio 12 compiler on IA32 due to unsafe FP comparison.

r1465: fix crashes with Sun Studio 11 compiler on Ultrasparc.

merge r1259 - add punch continuum to show fine - coarse continuua

merge back of several merges of mainline into 07.02 branch.
rev 1087 - set initial value to beta version in cdinit.cpp - now set to 0 as it should be in rev, but reset to 1 on this merge, which will become rc1 of 07.02

rev 1086, revise readme_docs.htm

rev 1082, 1084, rename docs

rev 1081, ionbal.h, parse_set.cpp, and zero.cpp, implement set atomic data sulpher DR command, and revised treatment of DR guesses in ion_recomb_Badnell.cpp

ion_recomb_Badnell.cpp had several conflicts where were resolved by hand

*.in merge back rev 1081 changes due to S DR

branches/c07.02_branch/source/lines_service.cpp: changes in HLineTransOpacSet to fix cache errors
trunk/source/lines_service.cpp: changes in HLineTransOpacSet to fix cache errors
branches/c07.02_branch/tsuite/auto/*: small changes in asserts as result of changes in HLineTransOpacSet
trunk/tsuite/auto/*: small changes in asserts as result of changes in HLineTransOpacSet

date.h - explain that day is now the subversion number, 00 for initial release

rest are merge of r919 on mainline

merge of r883

cdinit.cpp - lgRelease now true, this is a release version, so fixits will no longer print caution
path.cpp - now points to data within branch not trunk
runall.pl - uses release of branch

merge of r849

parse_puncch.cpp - BUGFIX logical flags for init clobber / no clobber of files was inverted

move PunchFilesInit from cdinit to cddrive

grains.cpp - upper limit on loop had MAX which consumed time - move to variable set before loop

grid_do.cpp - minor rm commented code

maincl.cpp - change filename to follow code naming convention

parse_crashdo.cpp - minor fix in comments

cddefines.h - add visual studio pragma to disable warning on use of str functions

ion_recomb_Badnell.cpp - cosmetic changes

*.in revise asserted quantities after bugfix in r828

tsuite/programs adjusted perl script run_programs.pl to work with cpp source and g++ - also now executes the program.  

cdinit.cpp - rm unused header, fix typos
cont_createpointers.cpp - typos, rm unused code
mole_co_etc.cpp - rm unused headers
mole_co_solve.cpp - rm unused headers
parse_punch.cpp - rm unused header, typos, fix format error in fprintf, in punch linelist command rm second file name early so that it does not trigger keyword,
service.cpp - typo

parse_punch.cpp invert logic for clobber file, now do not clobber by default, clobber is used to turn on clobbering files

h2_orion_hii_pdr.in and h2_pdr_leiden_v4.in - update close asserted values

these are a series of updates related to the Gu et al. 2006 UTA transition data
MAJOR CHANGE - bug fix in gu et al data set, do not use Kisielius data by default

UTA_Gu06.dat - this file was sent to me by Ming on the day that their paper was posted on astro-ph.  I had emailed asking for it.  Tests with Netzer showed that the data file was quite different from the version actually published in the ApJ.  The data file was updated to the published version but with integers for the roman numerals and an updated magic number

tests showed that the Kisielius UTA data set were very irregular and agreed poorly with Gu et al. data, which has smooth variation with charge.  Do not use Kisielius data by default any more - use Gu et al. 2006

source files changes are related to using this data file and including UTA line heating

the heat element of the UTALinies structure now contains the heating per UTA pump.  this is different from the usual meaning of heat in other instances of the EmLines structure.  heat is now set one time at setup and in several places protected against being set to zero (which is needed for usual EmLines data).  

atmdat_lines_setup.cpp - do not init heat to zero

atmdat_readin.cpp - update magic number for Gu data file, set heat as incident photon energy minus shell energy, set energyErg for UTALines (had not been set to anything before), define heat as heat per pump - Gu data set does not have upper and lower statistical weights defined, so use absorption oscillator strength to get Bul, define heat for all UTALines, not just new Gu data set,

ionbal.h - rename lgInnerShellRomas to lgInnerShell_Kisielius since paper is Kisielius et al., add **UTA_heat_rate to save heating per element and ion

conv_base.cpp - define UTA_heat_rate for all elements and ions

heat_punch.cpp - identify UTA heating as string when important

iso_create.cpp - do not zero UTALines heat since contains heat per pump
rt_tau_init.cpp - do not zero UTALines heat since contains heat per pump
rt_tau_reset.cpp - do not zero UTALines heat since contains heat per pump

zero.cpp - set up UTA heating array ionbal.UTA_heat_rate, do not use Kisielius data by default
heat_sum.cpp - define UTA heating as local var photo_heat_UTA, clean up comments

parse_dont.cpp - no uta command no longer has "romas" keyword - command is not only "no uta" with no options

parse_set - "set UTA" now has more options.  "set UTA Gu06" has keywords ON and OFF - default is ON.  "set UTA Kisielius" is new and has keywords ON and OFF and default is OFF - this is a change - had been used by default


*.in - these changed as result of changes in UTA treatment

revise nova_photos.in - Lya dest by FeII absorption drifted down

*.in - these changed as result of e2 bugfix r786

conv_base.cpp - only reevaluate UTA pump rates when lgRedoStatic is true, this will save some time

hydrobranch.cpp - init nd2 var to impossible value to prevent possibly use of uninit var warning

service.cpp - add change comment and date for PvH fix of e2 function.  This was a serious bug that had been in the code for some time, did change results, and worthy of a comment

nightly run tsuite - better load leveling
readme.txt - description of this

slow / h2_*.in files - update values after turning on McCall H3+ dr rate

assert_results.cpp - add missing DEBUG_ENTRY, EXIT
cpu.h, add pragma to disable warning on if constant expression
helike_einsta.cpp - minor formatting
zero.cpp - add cast to double to float

*.in - changes caused by update H3+ rate to McCall et al. nature paper

ism_grid.in - forgot to add in previous rev - this varies hydrogen density, shows solution going from warm ionized to cold atomic phases

h2_th85orion.in - asserted value changed due to CT chem reset error fix

cddrive.cpp - fix print formation when line not found

ion_nitro.cpp - small change in logic setting up trace flag in call to

ion_solver.cpp - - minor changes in print, add co chem ct recom and ionization to debug print - no change in routine's results

iter_startend.cpp - save and reset chemistry charge transfer that affect ionizatino of elements in ion_solver, for catalytic reactions

mole_co_drive.cpp - minor change in comments

mole_co_etc.cpp - detect if number of co species has increased after initial malloc, and stop.  this happens in tsuite/programs comp4 as reported in ticket 8

bugfix mole_h_step.cpp - change bug entered in R743 caused access violation - now do assert if ratei is correct value 

prt_comment.cpp - tidy print state - near lines 1424, but this print should not happen for levels within topoff - also cast bool to int for use in arith

reset chemical rates in iter_startend.cpp fixes ticket 7 in trac repsotiry, also causes changes in orion_hii_pdr.in - orion_hii_pdr_pp.in - and pdr_th85ism_cgto.in 

pn_paris.in - small drift in asserted values

helike_cs.cpp - comment out open mp call

parse_print.cpp - print size of char

func_test.in - print path

atmdat_2photon.cpp - introduced error in previous rev - logical flag was correct in r730, is correct now

two sims changed asserted continuum in ir as result of spline rather than linear interpolation, fix in r730

0.05 decrease in column density of H3+.  this had been true of newmole before the merge.  this sim tests the implementation of the Federman non-equilbrium hack, in which energy barriers for reactions are removed by turbulence.  the set non equilibrium chemistry command sets the var co.lgNonEquilChem which removes the energy barrier.

these are the changes in the slow test suite that resulted from the Badnell DR

tsuite/auto/run_parallel.pl:
  Fix some problems in the script:
  - delete optimal.in so that it doesn't fool checkall.pl
  - change "transread.out" -> "func_trans_read.out"

no botches but these three were fairly close.  test suite run cleanly

futher update of sims that changed due to Badnell DR

continue updating changed results as consequence of using Badnell DR rates

updating predicted results after Badnell DR enabled

these are the changed input scripts, due to the use of Badnell DR rates

this is less than half the changes.  more later.

atmdat_readin.cpp - add comment with data saying that changed from default behar to gu UTA data

*.in - these changes are in ionization of Fe due to differences between new Gu and older Behar data

zero.cpp - turn on Badnell radiative recombination by default

zerologic.cpp - bugfix flag for printing continuum indices was on by default, should have been off

sims changes as result of using Badnell RR rather than older values
most changes were due to changes in the Mg+ and Al++ rec - no dramatic changes resuled, but some larger changes were
coll_t5 mg 2
igm_lalpha - mg 1,2 balance
pn_paris - mg 2 2798 a lot
stars tests - al 3 fraction changed a bit, al 2 and 4 also

tsuite/auto/clean_tsuite.pl:
  new script to clean out the test suite after a test run.
tsuite/auto/run_parallel.pl:
  updated sequence of models in view of new execution times.

tfidle add alnte to phycon struc, the ln of the temperature
phycon.h - same

cool_oxyg.c - NA update [OIV] collision strength to Tayal 2006 - value is smaller so gas is warmer

atom_seq_boron - only changes in comments

atmdat_char_tran.c - use alnte rather than reevaluating

*.in - changes caused by decreased cooling in th [OIV] line

tsuite/auto/stars_optimize1.in:
  Fix test suite failure - TABLE STAR TLUSTY syntax should have been updated in r571.

source/parse_table.c:
source/stars.c:
source/stars.h:
  - Add support for new Tlusty BSTAR2006 grid; added keyword to TABLE STAR TLUSTY
    command to choose between OSTAR2002 and BSTAR2006 grids.
  - COMPILE STARS "userfile.ascii" now always calls ValidateGrid().
  - Some minor improvements.
tsuite/auto/stars_ostar2002.in:
tsuite/auto/stars_ostar2002_all.in:
tsuite/auto/stars_ostar2002_3d.in:
  - Updated command syntax for TABLE STAR TLUSTY command.
tsuite/auto/stars_bstar2006_3d.in
tsuite/auto/stars_bstar2006_all.in
tsuite/auto/stars_bstar2006.in
  - New test cases for the BSTAR2006 grid.

a wide range of trivial changes in response to compiling with /W4 on intel icl compiler, the windows version of icc.  many were static routines that had not been declared static.

cool_oxyg.c - update [OI] optical line collision strengths to Barklem,P.S.,2006,A&A (astroph 0609684) - collision strength for 6300,6363 double can be as much as 3x larger

parse_print.c - add some constants from float.h

*.in - update asserted line strengths - mainly effects of larger [OI] collision strength

Merging newgrids branch.

This branch had several goals:

1. restructure the code for handling stellar atmosphere grids.

  The main reason for restructuring the code was that there was a lot of code
  duplication, leading to lots of trivial errors. The restructuring has split
  off all duplicated code into separate routines. There were also minor
  differences in the way each grid was set up, making them incompatible. These
  differences have been removed. Now the new structure is in place, adding new
  grids will be very easy. The code has been rewritten to handle grids with
  more than 2 dimensions. This enables interpolation in metallicity and
  possibly other parameters (e.g. Mdot).

2. updating the format of the ascii and binary atmosphere files.

  Since the changes above did break all existing binary atmosphere files,
  this was a good moment to update the format of both the ascii and binary
  files. The most important changes were that the ascii files got version
  numbers, and that the binary files for 32-bit and 64-bit versions of the
  code are now compatible.

3. add support for new grids, and update existing ones.

  - the Kurucz Atlas models have been updated to the most recent version,
    and support has been added for non-solar metallicities as well as the
    ODFNEW models.
  - The old Kurucz (1979) and Mihalas grids are no longer hardwired into
    the code, but are now split off and treated like any other grid.
  - Starburst99 output files are now transformed into the same format as
    the atmosphere grids. The TABLE STARBURST command has been retired.
  - the Rauch H-Ca grid has been updated to the final version.
  - 3-dim versions of the Rauch H-Ca and H-Ni grids have been added to
    enable interpolation in log(Z).
  - support has been added for the Rauch pure hydrogen, pure helium, H+He,
    and PG1159 grids.
  - support has been added for the Tlusty OSTAR2002 grid (all metallicities).
    The TABLE TLUSTY command has been retired.
  - support has been added for the WMBASIC grid of O-star models.

source/stars_atlas.c:
source/stars_costar.c:
source/stars_kurucz79.c:
source/stars_mihals.c:
source/stars_rauch02.c:
source/stars_rauch_old.c:
source/stars_rebinatmosphere.c:
source/stars_werner.c:
  Deleted and merged into
source/stars.c

source/stars.h:
source/cont_ffun.c:
source/parse_table.c:
source/parse_compile.c:
  Modified to support new grid structure.

tsuite/auto/stars_costar.in:
  Renamed to
tsuite/auto/stars_costar1.in

tsuite/auto/stars_kurucz.in:
  Renamed to
tsuite/auto/stars_kurucz79.in

tsuite/auto/stars_rauch.in:
  Renamed to
tsuite/auto/stars_rauch_h-ni.in

tsuite/auto/stars_rauchold.in:
  Renamed to
tsuite/auto/stars_rauch_h-ca.in

tsuite/auto/stars_mihalas.in:
tsuite/auto/stars_starburst99.in:
  Minor text changes.

tsuite/auto/stars_starburst99.dat:
  Deleted.

tsuite/auto/stars_wmbasic.in:
tsuite/auto/stars_atlas_3d.in:
tsuite/auto/stars_ostar2002.in:
tsuite/auto/stars_costar4.in:
tsuite/auto/stars_optimize1.in:
tsuite/auto/stars_rauch_3d.in:
tsuite/auto/stars_rauch_hydr.in:
tsuite/auto/stars_rauch_h+he.in:
tsuite/auto/stars_ostar2002_all.in:
tsuite/auto/stars_optimize2.in:
tsuite/auto/stars_rauch_helium.in:
tsuite/auto/stars_ostar2002_3d.in:
tsuite/auto/stars_atlas_all.in:
tsuite/auto/stars_rauch_pg1159.in:
tsuite/auto/stars_optimize3.in:
tsuite/auto/stars_atlas_odfnew_all.in:
tsuite/auto/stars_costar2.in:
tsuite/auto/stars_costar3.in:
  New test suite cases to test new grids.

source/mole_co_reactions.c
source/mole_co_step.c
  Fixed svn:eol-style property.

changes in asserted values in test suite from rjrw merge of co chem network

mole_co_drive.c - near line 475, add test on the H+/H ratio, and do not do the co network if H+/H>0.5.  this was added by NA to delay evaluation of the co network until conditions become appropriate.  It had little effect on the test suite results but should speed up the code.

orion_hii_pdr.in - - 0.1 dex changes in intensities of very optically thick CO lines as result of delaying evaluation of CO network until later.  this is down in the noise.  changed values and increased uncertainty in asserted co lines to 0.15 dex, the uncertainty in the column densities.

grid routines altered by Ryan Porter so that punch does not output header for each grid point and third parameter on grid command is increment, as in a for loop, rather than the number of points

parse_punch changed to write header into string in all instances, then decide at end of routine whether to write the string, to implement not putting headers on each grid point

punch grid command introduced

punch options in func_grid.in changed to 

Ryan porter change more instances of iso.numLevels_max to iso.numLevels_local for continuum lowering

scripts - load leveling - lightning now slower, do fewer there, more on cumulus

LineList_PDR_H2.dat - att higher H2 lines needed for Helix project

func_grid.in - order of options on grid command changed

these are a series of changes related to the grid command, and increasing the number of comments in the he-like code.  All changes to the source were made by Ryan Porter.

the grid command now requires three args, the lower and upper bounds, and the number of points.

in the orion_hii_pdr_xx models two changes - total pressure with magnetic field, and stop column density 22.  first is more realistic, second is because most molecular clouds have similar column densities, about long N 22.  these two had gone to column densities of 25.

lines_service - add routine ConvRate2CS to convert line deexcitation rate into collision strength

cool_oxyg update to Krems et al. H0 collision rate for [OI] emission

he1n2t4.in - wavelength of 10830 slightly different - this uses more sig fig for lines

update results from slow runs - these were largely changes that were done in Meudon, changes in [OII] cs and Bowen efficiency.

small changes in comments and print statements

cool_oxyg.c - HN improve fit to OIII cs, better temp dependence

ion_recomb_Badnell.c - update Fe DR and RR to his astro post of aug 2006, does Fe8+ through Fe14+

change in fit to OIII cs and much larger fe dr changed several simulations

load leveling in night run scripts/nightly_run_tsuite.pl
update slow sims to catch up with past few weeks of changes

finish previous commit - versions had gotten out of synch - the completes charge transfer revisions

this undid the changes introduced by the Mg0 and Si0 charge transfer.  the umist rates are totally bogus

cddefines.h - redefine macros for electrons, protons, and he+

cloudy.h, iterations.h - move defn of iter_end_check from cloudy to iterations 

cont_setintensity.c - print warning if incident continuum is surprisingly faint, also do not include radio continuum in this test since cmb itself could be significant

conv_init_solution.c - when ices are important do not let temperature change by much - large T changes with ices throws the solvers badly off

cool_iron.c, grains.c, ion_solver.c, iter_end_check.c, parse_commands.c - minor changes in comments or print statements

zero.c - init radiius.drNext to 1

func_trans_punch.in - last became missing on punch transmitted continuum - no idea how this happened

this runs test suite cleanly

atmdat_char_tran.c - update mg0 & si0 charge trans ion with H+ to umist 05

atom_pop3 - assert physical quantities are >= 0

cool_oxygen - update [oii] and [oiii] collision strengths

grains.c - attempt at getting bakes hack to work with grain physics is turned off - this is critical for debugging hydro models.  heating now vastly too large and commented out

prt_final update [oiii] cs to be parallel with cool_oxygen


*.in - many changes in results due to CT and cs changes

bands_continuum.dat - NA add total lumin

this updates the auto test suite to incorporate results of two major changes.

in ref 446 the hydrogengic collision rates were changed to vriens & smeets by Ryan Porter.  These were many dex smaller than had been found by Ferguson's incorporation of another approximation.  This mainly affect optical lines from very dense BLR clouds - the lines are generally stronger, often by factor of two.

next difference is in the charge transfer rates introduced in rev 447.  this is a major ionization source for Mg and Si.  PDR models changed as a result.

with this set of changes the auto suite runs cleanly 

slow does not

the DR rates in conv_base.c have been updated over the past few weeks to preprints and pre-publication values computed by Badnell & Gu.  these are not used by default and so have had no effect on the test suite.  however they are used in establishing a mean dr for ions that have no rates.  with this change the new dr rates are allowed to enter in the mean DR used for guesses.  

the test suite sims that changed tested ions that used these fake DR rates, so these changes represent real uncertainties.  

update H - N charge transfer, also Mg IV collision strength

atmdat.h - add HCharExcRecTo_N0_2D, rate coef for H0 +N+ -> N0* charge transfer into excited state producing NI 5200

atmdat_char_tran.c - add Yun fits to Stancil H0+ - N0+ charge transfer

cool_nitr.c - minor change in comments 

prt_lines_lv1_li_ne.c - add H - N charge transfer excitation of NI 5200, in total NI line intensity, include CT and chem production - according to stat weight - not important in test suite

cool_magn - update mg iv 4.487 micron line collision strength

*.in - update predictions in test suite for sims that changed as result of updated H-N charge transfer rate

cool_neon.c - update collision strengths for NeII - NeV

mole_h_stop.c - comment change 

parse_crashdo.c - print statement had invalid argument - fixed

atmdat_char_tran.c - add commented out block of code for Terry Yun

*.in - updated asserted values that changed as result of change in neon collision strengths

stars*.in - added constant temperature command to all since there test shape of ionizing continuum, not thermal physics

Mitchell Martin add Fe+13 - Fe+12 DR rates, at this time do not add to mean for
	dr since this changes mean significantly.  will update when Gu & Badnell
	numbers are also in

timesc.AgeCOMoleDest is now a vector for all CO species, keeps destruction timescale for all

bugfix - punch_do - every option was added to punch continuum - did not work at all,
	punch continuum punch first zone only.  fixed.

PAF complete update headers for doxygen

runbotched.pl and runcrashed.pl - rm nice from system run

conv_init_solution.c - on iterations after first, set initial temps, other physical variables, to previous iteration's converged temp in first zone, had always used init guess of first soln

dynamics.c - set const temp to 0 to turn off coronal equili command after relax, in time dependent models

mole_h2_etc.c - fix bug in h2 coll interpolation command found by Mitchell Martin

runall.pl - rm nice command for run all

parse_cosmic_ray - add cosmic rays equipartiton option to cosmic ray command

hextra.h 
define CR_EDEN_GAL_BACK_EV_CMM3 as Webber background cr energy density
cr_energydensity added
lg_CR_B_equipartition set true if equipartition

highen.c - if lg_CR_B_equipartition set true then derive cr energy density and ionization rate from B field

physconst.h small changes in some comments.

pressure_total - add todo to add cr pressure near line 204

zerologic - set some B equipartition variables to zero

orion_hii_pdr_pp.in - add punch temperature to get te & deriv

assert_results - add three molecules to those whose column densities can be assertetd
cddrive.c - add three molecules to cdColm 
three added are CN, CH, and CH+

mole_co_etc.c - gather together rates that depend only on temp or rad field, and do not reevaluate them unless quantities have changed - several duplicate rates would found and removed, and two cases where rate multiply defined with second being incorrect were found.  this caused some changes in molecular column densities - not large

prt_lines_lv1_k_zn - make list of Fe 3 lines that come from model atom added by Blagrave


perl scripts - remove nice -n 5 since this pushed scripted runs to too low a priority

*.in - small effects of bug fixes in mole_co_etc.c

pdr_co_fully_noneq.in - add this with non-equil chem turned on
pdr_co_fully.in - add turbulence to this so that pdr_co_fully_noneq.in is totally parallel

major refactoring of co chem files:
co chemistry refactored by Nick Abel to use global rate coefficients which are evaluated only in one place, and used elsewhere.  this uncovered bugs where the same rate was not used in different routines. 

orion_hii_pdr, orion_hii_pdr_pp, pdr_co_fully, and pdr_leiden_?3.in changed as a result of bug fix of inconsistent rates

moved nitrogen cooling variables out of coolheavy into nitro.h

hmole_reactions now called from CO_update_chem_rates in mole_co_etc.c

structure mole_co_priv created to hold rate coefficients in global name space

add files mole_co_priv.h and nitro.h

NI 5200 recombination from Pequitnot et al.  1991 (must fix reference) prt lines lv1 li ne line 302

rm optimal.in from repo since this is generated by optim*.in scripts during nightly testing

weekly_backup.pl - change target drive to cumulus raid

svnbackup.pl - target changed to cumulus raid

RP change wavelengths of various He1 lines to greater precision

feii_blr* add punch feii lines

add test of optimize grid to test suite

RT update lines_service.c so that h-like As for levels n > 15, scale with As for lower levels.  effects are important for fit to case b as per ferguson & ferland 97 - fixed discontinuity at higher levels, n >15, where As were not adjusted for density

*.in, small changes in asserted higher n lines at ~5% level

perl scripts, rename current to trunk

pressure.h, struc.h - change name PresRadCurr to pres_radiation_lines_curr to make clear that this is line radiation pressure not continuum

punch pressure, column title Prad changed to Prad(line) to make clear line not continuum

parse_atomhlike.c atom h-like pyman pump scale factor will be log if negative

add pressure to dynamics*.in

*.ini, change "print on" to "print on " - gcc on two new machines did not match on keyword - for some reason trailing space not added on with gcc

orion_hii_open.in - at print line optical depths to see HeI optical depths in orion

prt_alltau.c - changes in print helium optical depths - give lowest give continuum edges also give several explicit lines rather than cryptic code

merge Ryan porter's changes to both helike and grids code.  many were in response to lint.  

change DR for helike species to Badnell DR rates.  the set recom command does not change this.  This update changed results for species with H-like, He-like abundances, so a handful of *.in changed

conv_fail.c - give more useful message if code fails because we went into molecular gas without cosmic rays

ion_trim.c - identify case where hden has become very very small and floating abundances have underflowed below zero.  abort in this case.  previously had detected insanity with an assert.  now says to consider turning off the element or not running such low densities.

parse_crashdo.c - small change to comments

punch_line.c add tabs between numbers of punch line cumulative

feii*.in add punch feii level populations

h2*.in slow run - changes introduced by Nick Abel adding N2H+

NA break mole_co_step into two pieces, _1 and _2, to speed up compilation
remove mole_co_step

NA add N2H+ to chemistry

iter_startend better doc of problem with stop line command

some test cases had changes in SiO and H3+ as a result - botch but not big botch

scripts changed to use gcc for day runs on lightning

remove unused header files from several source files

rt_line_all.c, rt_line_one.c - reevaluate line rt if ionization trim occurred

func_ion_increase.in - do not assert frac abun of highest fe stage since only forms in five zones, depends on zoning

radius_next add logic to look at previous iteration on time dep recom to get dr

nlr_paris.in - bug in NI changed predicted ni int

revise optimized models with new [OII] As - reran at n=5 and used that spectrum as the one we are after.

pn_paris.in - [OII] revised

these are all trickle down effects of change in [OII] As

include turb press with flag to turn off in punch output

bugfix - correct interaction between turbulent equipartition and set pressure 

slow sim results updated - these changed as result of bug fixes that went also into 06.02a

hazy3_06_02.pdf - change opening index so pages are collapsed

parse_atomhlike.c - do not allow fewer than 5 levels so that Hb can be referenced

parse_drivecmnd.c - delete drive helium command which no longer worked

prt_comment.c - small change in format of prints

limit_veryveryfast.in - cleaned up atom levels commands, do not request 4 levels - imposed limit is now 5 to include Hbeta

all changes are in computing the emergent continuum

old dst lin array removed and main sumlin made vector with two elements, first intrinsic spectrum second emergent.

print line emergent tells cdLine to get emergent rather than intrinsic

bugfix - typo in corrected value of asserted quantity in this sim

previously had added C+, Si+, and S+ ct with Mg0 and Fe0, and bug fix since these were not evaluated when CO chem was not done.  ct is now in and C+ and Si+ abundances are now much smaller.

all minor changes in comments or spellings - no source code changes

add comments to atom_pop5.c and sample call to cool_nitr.c

add two sims to test suite, func_t10.in and func_t30.in, that exercise the code at 3K and at 1e10K, the low and high temperature extreme where the code is designed to function.  They exposed several bugs which were fixed in rev 197 and 198, and merged onto c06_02_branch at 199

added series of agn models to test temperature in inflection point of Spitzer S curve for three-phase AGN cloud stability.  Temperatures around U = 2.25 depend quite strongly on the DR rates for 10 and fewer electron ions of iron.

minor adjustments to asserted values of other sims in auto part of test suite

added S curve figure to hazy_kmt sims

atmdat_dielrec_fe.c - add comments and reference

cdinit.c - say that compiled date is date cdinit.c was compiled

ion_recomb.c - reorganize logic to be simpler & clearer in kludge and low T dr

parse_set.c - no longer a steve option (was undocumented) but brought noise option out as a diel rec option

return to lexington after two weeks in Pune.  Most ports were blocked so it was not possible to update repo for entire time.

major activity:
Ryan Porter updated iso levels to include continuum lowering for hydrogen like series.  not actually turned on but it could be.

xi command, parse of ionization parameter in parse_ionpar.c started in col 5 so often would miss start of number

printed H2 column density at end of standard output was total column density in protons not H2 molecules.  changed to total number of molecules.  only one print statement was affected.

add thermal destruction of ices for kinetic temperatures greater than 100K

changed CO ice on grains in h2_cr_grains.in and h2_solomon.in

update comments in bands_continuum.dat

slow run changes resulting from missing large block of code

atmdat_char_tran.c - adjust format of atomic data refence - comment only
doc_* generated artifacts

prt_lines.c - add assert at end of routine to confirm that number of lines did not change from initial calls (ipass==0) where labels and wavelengths are set up, and later calls (ipass>0) where line intensities are added to stack.

adjust title of four blr sims

source/cdinit.c:
  Fixed broken compiler detection for icc, as well as gcc on SGI MIPS.
  Added FPE trap support for HP compiler on sdx.
tsuite/auto/run_parallel.pl:
  Improved cleanup (leave all files from repository) and sequence of sims
  (make sure the long running ones are upfront).

scripts - cosmetic changes

mean.c fix typo at line 166

programs - change calls to cdGet that reference routines with changed names

update badnell_rr.dat to nigel's current version
edits in hot_issues.txt and to_do.mdb
collion now use Badnell DR data

further clean up of programs html and supporting graphics files.

added comp4.htm to repo

minor additions to repos of file missed last time

changes to program docs to make better describe their purpose. 

two independent changes here:

the uta directory below data should not have been included in the repository.  they were working files that were provided by Kisielius & Keenan.  Deleted.

Merged in Ryan Porters changes to the He-like iso sequence.  these made a finer grid on the tabulated recombination coefficients, used a higher order interpolation method for extracting these recombination coefficients, and adjusted the test suite to trace the changed asserts.  

set line precision command added by RP - increases number of digits in printed wavelength.

cdH2_Line was added to extract a particular H2 line from its quantum indices.

dielectronic recombination command changed to set direlectronic recombination

RP made changes to optimization code used to create xspec interface.

believe that test suite will run cleanly.

these are a series of changes relating to multi-level models of Fe emission.  new multi-level models of FeIII, FeXI, and Fe XIII were introduced.  other changes were to infrastructure in support of these new models.  Some atomic data were updated, and other FeXI lines removed, from level1.dat.

correction to utilitymacros.h fixed warning issued by g++ when *.c renamed to *.cpp

agn_warm_absorber.in and blr_n09_p22_Z20.in had asserted results that changed as result of improved Fe emission.  Updated to current predictions.

h2_orion_hii_pdr.in - failures in CO chemistry were due to going so deep into the molecular cloud that grain temps fell to such small values that catalysis of H2 no longer occurred.  Cosmic rays still destroyed H2 so H had a transition from H2 to H0.  changed to stop at AV of 1000

tsuite/auto/run?.pl - path on std has remained current not trunk, adjusted path in these six scripts, which are used on the sdx

tsuite/programs - perl script run_programs.pl updated to link into gcc object files

small edits to all programs

cool_iron.c - added kevin Blagrave model of FeIII atom.  
prt_lines_lv1_k_zn.c - This adds many FeIII lines.  several revs before FeII cooling was changed to Fe2c in output.  this rev change Fe III and Fe IV cooling to Fe3c and Fe4c by analogy.

blr_hizqso.in - added FeIII cooling changed intensities of two lines.

ion_solver.c - when ices form deep in cloud the molecule solver can take many iterations to settle down.  During early stages the molecular network can be out of synch with the ion solver and the atomic density, the difference between the gas phase and molecule density, could be negative.  ion_solver.c was aborting when this happens, but tests show that the network and ionization ladder do eventually settle onto a good solution. chng is to now print a PROBLEM statement when this occurs, and set a non-convergence flag, but soldier on.

zero.c - the limit to the ice fraction is no longer needed due to above - change is to set ice fraction for stopping to >1 so will not occur.

prt_lines_lv1_li_ne.c - bug in [NII] 5755 recombination contribution reported and fixed by Kevin Blagrave.

several small formatting changes in comments

copy_tsuite.pl now does both auto and slow dirs off tsuite 
it still needs to do the same programs

the h2_pdr_leiden sims had not been updated since NA increased the chem network to include molecules.  one sime, h2_orion_hii_pdr crashes - the only major problem with the test suite.  with these changes the auto will go cleanly as will the modified slow scripts.

in zero.c made Torsten Elwert's approximation to the Solomon rates the default.  pdr_leiden_hack.ini - added command to make BD96 used for the hacked models.  many pdr models changed as a result, documented and changed the asserted values.

radius_next.c - add logic to follow changes in ices when they become more abundant that 1e-5 of main element.  ices form exponentially when grain temperature falls near the sublimation point.

This exposed a bug in the FeII optical depth routines.  atom_feii.c used the number of levels in the atom to reset optical depths, but this has been decreased in the orion_hii_pdr_cc sim since it went so deep in the molecular cloud.  The full range of levels is now reset.

several variable names were changed to follow the style of the code.

tsuite/auto/run_feii.pl
  fix svn:eol-style and svn:executable properties.

end a series of changes dealing with H2 photo heating

heat_sum add photoelectric and bound compton heating by H2, H2+, and H3+

mole_h_step usee Yun et al. photoionization cross section for H2, for H2+ had used 2*H0 photo rate, change to just H0 rate since this is a one-electron system

opacity_addtotal.c use Yun et al photo cs for H2 opacity, had been 2H0, include H2 in bound compton opacity, had not been included

work on H2 photo cross sections - photo rate and opacity in but not yet heating

partially add Dalgarno cs for H2 photoionization - only in for rate, not opacity or heating

h2_t500 results had not been modified to results after NA include molecule freeze out

Changed eol-style property to native on various *.dat files in data/uta and
  tsuite/auto.

optimize_input.c: bug-fix
  variable limit exceeded was handled incorrectly by optimizer.
grid_input.c: bug-fix
  same problem, same fix.

adjust file properties for ascii and perl files - no source change

minor changes to input scripts

update test suite file punch output to have similar second parts of names