By Jixie Zhang @20180306 This program is called STEG. It is used to generate xs map and generate events for simulation. So far it is dedicated for EG4 experiment in Jlab.

20180416 I am going to modify STEG to bin data in SinTheta other than CosTheta. The modification on April 9th is the last version that bins in CosTheta. I created a branch named 'BinInECosTh' for it in github.
From now on, the master branch bins in SinTheta.

04/16/2019

1. Modify radLengthEG4_simp.f to support alternative PF values.

04/02/2019

1. Change variable 'UseLongTarget' to 'defaultTarget' in kpaVarChanges.inc. Modify the code to support top, short and bottom target cells.

03/18/2019

1. Modify Structure function choice 22. For SF=22, only allow R=R+dR. before it could be R=R-dR or R=R+dR.

02/07/2019

1. Modify newSFs.f to make A1=A1+changeA1 take effect for w>=2.02
2. Add variable 'UseLongTarget' into kpaVarChanges.inc to support short target cell.

01/08/2019 This version has a big change comparing to previous one.

1. New structure function subroutines. Modify several files.
2. Add newSFs.f. Change it to make A1=A1+changeA1 take effect only for w<2.02

04/16/2018 This version has a big change comparing to previous one. All tools have been rewritten. 0. Place a copy of 20180409 version as a new branch in github, named 'BinInECosTh'

1. Binning in sinTheta instead of cosTheta.
2. Formated some screen outputs

04/09/2018

1. Added "eloss_ion_prob.f" to calculate most probable ionization energy loss.
2. Modified stegAll_myOwnNoAvgXSinBinMultSc.f to call eloss_ion_prob(XXX): A)Need to call SET_THINGS_UP(XXX) in both creating xs map mode and throw event mode. From now on input file "rcslacpol.file" always required! (this file defined the target type.) B)Added p0_e and th0_e into output ntuple
3. Changed some print out message in file read_mapMyOwnVersion.f and set_things_up.f

03/19/2018

1. Fixed typo in Makefile. very critical!
2. Changed stegAll_myOwnNoAvgXSinBinMultSc.f to make it compatible with Lamiaa's version. If file 'stegAllvPREG.dat' exists in currrent dir, will read arguments from this file. Otherwise, ir will read arguments from shell (usually run in this way: stegAllvKPA_rhel7 < inputfile )
3. Changed stegAll_myOwnNoAvgXSinBinMultSc.f. During generating events, print event# message every 1000 events. Before it prints this message every event.
4. Fixed bug in file nThnPnSum.inc. (old) parameter(num1Dbins=n_p_eln_p_el) (new) parameter(num1Dbins=n_p_eln_th_el) This bug does not do anything bad in the past since n_th_el is equal to n_p_el. But it will if n_th_el is not equal to n_p_el.

03/15/2018

1. Solved bug about signning EB_INDEX.
2. Files newSFsSEK17.f sek_meKPA.f sigmarNEW_SEKnw17.f are in CRLF(windows) ending format, Changed them to CR(Unix) ending format. Changed all of them to EMACS indent format.
3. Replaced YoniSEK.f using the original version without my format.
4. Added file Changes_Since_V1.0.txt make it possible to trace back to version 1.0.

1. Found that the xs under elastic peak seems abnormal large than its neighbour points. This is due to the fact that in some theta bin, the center value of the theta-p bin is too close to the E'_elastic. Since we are not simulate elastic peak in STEG, we would like to remove these points. The following cuts are applied:

   XCUT < A*0.99; E' < E'_elastic -E'_bin_width; W > 0.95 .

2. During throw event, called Xiaochao's subroutine 'rleg4_simple(xxx)' to calculate TA,TB. TA will be used to determine MSC for theta of the scattered electron.

03/06/2018 Reorganzied this program and put it into github.

1. Essential files for successful compilation are all in src: *.f, *.inc
2. All essential file for running this program are in otherfiles: F2ALLM07.inp?, rcslacpol.file, *.ful, *.dat
3. Makefile changed

02/28/2018:

1. Added Xiaochao's subroutine 'rleg4_simple(xxx)' to calculate TA,TB,PF according to given beam energy for NH3.

2. Changed set_things_up(xxx) to sign EB_INDEX according to given beam energy and target type.

3. Removed MSC in phi angle when throw events.

4. Changed XCUT to A0.99D0. Previously it was 'XCUT = 1.0/(0.0658/(EEP*sin2)+1.0)'.

5. Cleaned up some useless comments, removed some unused subroutine.

   TODO: clean up, reorganize, re-format then upload into github

02/06/2018

1. STEG outputs 'infinite' or 'NAN' occasionally

According to fortran language manual, IEEE 754 floating point numbers can represent positive or negative infinity, and NaN (not a number). These three values arise from calculations whose result is undefined or cannot be represented accurately. They are defined as 1/0 = ∞ log (0) = -∞ sqrt (-1) = NaN

I looked into the source code and the output. I found that these issues occur only at the edge of the kinematics acceptance, for example, E'>E_elas due to round off effect or due to not precise input value of Proton|Deuteron mass.

Feb 06: A subroutine to check infinity or NAN is added. if any of xsB, xsR, xsP and xsPR is 
        infinite or NaN, set xsB=xsR=xsP=xsPR=0.0.
Feb 06: solved!

2. STEG produces negative radiated unpolarized cross section, but born cross section are all positive or zero. It happens only to NH3 target.

   Feb 06: Added code to check the radiated unpolarized xs, if negative, set xsB=xsR=xsP=xsPR=0.0. Feb 06: solved!

10/13/2017: Debug for create xs map for NH3 target, use Sebastian's new code which can alter FakeProton between TRUE and FALSE.

Sep. 2017 Fixed bug in writing bos outpt

Sep. 2017 Installed CLAS package in CentOS7. Adapted both STEG and CombineFracMaps programs into my own environment.

June 2017 Copied Krishna's version

8/14/16:

Choose proper values of the four parameters/flags (changeA1, AsymChoiceKpa, SFchoiceKpa, and factorRA) in kpVarChanges.inc depending on which simulation we want to perform. Choose n_th_el=2, and n_p_el=200 etc in the file binning.inc to make exe for XS-map generation Choose n_th_el=n_p_el=200 or something like that for exe to use for event sampling/generation