Feb 12 2014   Wed Feb 12 21:20:48 2014 (gmt) (Stv+StarRoot)
=======================================
                    CVS ci good version Related testing in
                    http://www.star.bnl.gov/~perev/qa/STV_FEB_11_BaseQA.html
                    Used old version of StMultiKeyMap
                    Propagate by THelixTrack

Feb 17 2014   Compare versions with & without THelix. Used one file with 50 evts.
==================================================
                    File=rcf13003_77_50evts.MuDst.root
                    Results for NoHelix are slightly worse
Helix ON   
InputFile = /star/rcf/prodlog/eval13/log_Stv_Feb_11/trs/rcf13003_77_50evts.log
    nFiles =1 nEvents = 50 <nEvents>=50
    nGlob =      40673 <nGlob> =   813 nGlobMax= 3241
    nGgoo =      31632 <nGgoo> =   632 nGgooMax= 2441 Ggoo/Gtot=0.778
    nPrim =      20469 <nPrim> =   409 nPrimMax= 1448 Ptot/Gtot=0.503
    nPgoo =      19934 <nPgoo> =   398 nPgooMax= 1429 Pgoo/Ggoo=0.630

Helix Off
    nFiles =1 nEvents = 50 <nEvents>=50
    nGlob =      40659 <nGlob> =   813 nGlobMax= 3243
    nGgoo =      31589 <nGgoo> =   631 nGgooMax= 2435 Ggoo/Gtot=0.777
    nPrim =      20464 <nPrim> =   409 nPrimMax= 1448 Ptot/Gtot=0.503
    nPgoo =      19925 <nPgoo> =   398 nPgooMax= 1427 Pgoo/Ggoo=0.631

    Looking on Xi2 for primaries with low Pt HelixOn is slightly better again
    I am going to check accounting of energy loss
Helix Off + few cleanup
nFiles =1 nEvents = 50 <nEvents>=50
nGlob =      40655 <nGlob> =   813 nGlobMax= 3240
nGgoo =      31586 <nGgoo> =   631 nGgooMax= 2435 Ggoo/Gtot=0.777
nPrim =      20466 <nPrim> =   409 nPrimMax= 1448 Ptot/Gtot=0.503
nPgoo =      19927 <nPgoo> =   398 nPgooMax= 1427 Pgoo/Ggoo=0.631

Energy loss was tested by numerical integration. Result the same.

Feb 18 2014   Run 1094 files THelixOff.
==================================================
StiTime = 1.14

In File(s) /star/rcf/prodlog/eval13/log_Stv_Feb_18/trs
StvTime = 4.01
nFiles =1094 nEvents = 54600 <nEvents>=49
nGlob =   41127629 <nGlob> =   753 nGlobMax= 3617
nGgoo =   32091481 <nGgoo> =   587 nGgooMax= 2870 Ggoo/Gtot=0.780
nPrim =   21079767 <nPrim> =   386 nPrimMax= 1752 Ptot/Gtot=0.513
nPgoo =   20509561 <nPgoo> =   375 nPgooMax= 1722 Pgoo/Ggoo=0.639

The previous Feb11 THelixOn:
StvTime = 4.16
nFiles =1094 nEvents = 54600 <nEvents>=49
nGlob =   41140227 <nGlob> =   753 nGlobMax= 3615
nGgoo =   32132991 <nGgoo> =   588 nGgooMax= 2874 Ggoo/Gtot=0.781
nPrim =   21082626 <nPrim> =   386 nPrimMax= 1737 Ptot/Gtot=0.512
nPgoo =   20513758 <nPgoo> =   375 nPgooMax= 1707 Pgoo/Ggoo=0.638

Result is almost the same. Speed for HelixOff is slightly better (4%)
Looking on Fig12 Xi2 for small Pt is better then HelixOn

CVS Version Wed Feb 18 21:20:48 2014 (gmt) (Stv+StarRoot)
----------------------------------------------------------------------
This version is GOOD.
GoodOldStvTime = 10 (Jan 30 2014)

But Jason found that 1/pt:pt is much worse than Sti and worse than Feb11 HelixOn
I found that in versiom helixOff  was used  eLoss and then calculated pLoss/p
and in versiom helixOn  was used  pLoss/p was calculated inside of ELossTrak.
Which is completely wrong.
This was fixed and test was made:

In File(s) /star/rcf/prodlog/eval13/log_Stv_Feb_24/trs
nFiles =1094 nEvents = 54385 <nEvents>=49
nGlob =   40947304 <nGlob> =   752 nGlobMax= 3616
nGgoo =   31953351 <nGgoo> =   587 nGgooMax= 2871 Ggoo/Gtot=0.780
nPrim =   20991169 <nPrim> =   385 nPrimMax= 1752 Ptot/Gtot=0.513
nPgoo =   20423302 <nPgoo> =   375 nPgooMax= 1722 Pgoo/Ggoo=0.639

StvTime = 3.94
Now 1/pt:pt is the same as Feb11 HelixOn
But Still worse than Sti

It should be investigated.
===================
The possible reason: 
Geant move track along broken line. Stv broken line is more smoof.
I suppress smoofing in Stv and Leve run jobs

26.Feb.2014
http://www.star.bnl.gov/~perev/qa/STV_FEB_25_BaseQA.html#Fig.46
q/pt becames slightly worse. So the guess above iw wrong(as expected)

Hitless ends of track removed before refit., Then after refit Dca Node added.
Supposedly could improve Q/pt
http://www.star.bnl.gov/~perev/qa/STV_FEB_26_BaseQA.html#Fig.46
No improvement
nGlob =   40892585 <nGlob> =   752 nGlobMax= 3623
nGgoo =   31434607 <nGgoo> =   578 nGgooMax= 2817 Ggoo/Gtot=0.769
nPrim =   20871268 <nPrim> =   384 nPrimMax= 1716 Ptot/Gtot=0.510
nPgoo =   20181456 <nPgoo> =   371 nPgooMax= 1673 Pgoo/Ggoo=0.642

StvTime =3.97

Now Many Many Many Many Many ManyMany ManyMany ManyMany ManyMany Many
===============================================================
Feb27
Jason found MANY in TPSS which is veryVery bad.
Remove it, and:
In File(s) /star/rcf/prodlog/eval13/log_Stv_Feb_27/trs
nFiles =1094 nEvents = 54453 <nEvents>=49
nGlob  =   41025941 <nGlob> =   753 nGlobMax= 3606
nGgoo =   31581921 <nGgoo> =   579 nGgooMax= 2836 Ggoo/Gtot=0.770
nPrim  =   20906816 <nPrim>  =   383 nPrimMax= 1713 Ptot/Gtot=0.510
nPgoo =   20212271 <nPgoo>  =   371 nPgooMax= 1663 Pgoo/Ggoo=0.640

StvTime = 3.125
============

So Time is 20% better and Numbers are he same.
Goooooooooooooooooooooooooooooooooooooood

28.Feb.2014
Run with dP in ResetELoss very small (1e-4)
Result is slightly better and time is the same. (3.128)

================================================================
03/10/2014

New StTGeoHitShape introduced (Rmin bins added)
It should avoid tracking in non hitted area. Could improve time.
StvTime = 5.19 
nFiles =1 nEvents = 50 <nEvents>=50
nGlob =      40328 <nGlob> =   806 nGlobMax= 3248
nGgoo =      30834 <nGgoo> =   616 nGgooMax= 2361 Ggoo/Gtot=0.765
nPrim =      20218 <nPrim> =   404 nPrimMax= 1429 Ptot/Gtot=0.501
nPgoo =      19537 <nPgoo> =   390 nPgooMax= 1388 Pgoo/Ggoo=0.634

  Big Run result: Time 3.084 (1.4% better)
nFiles =1094 nEvents = 54510 <nEvents>=49
nGlob =   40708330 <nGlob> =   746 nGlobMax= 3580
nGgoo =   31267952 <nGgoo> =   573 nGgooMax= 2802 Ggoo/Gtot=0.768
nPrim =   20811553 <nPrim> =   381 nPrimMax= 1716 Ptot/Gtot=0.511
nPgoo =   20058245 <nPgoo> =   367 nPgooMax= 1654 Pgoo/Ggoo=0.641

Comparing with the previous (Feb27) result is slightly worse
Probably related to the direct limits which suppress some loopers
I will add some safe distance into StTGeoHitShape

10% safe area added
Result for 50 events:
nFiles =1 nEvents = 50 <nEvents>=50
nGlob =      40694 <nGlob> =   813 nGlobMax= 3258
nGgoo =      31176 <nGgoo> =   623 nGgooMax= 2385 Ggoo/Gtot=0.766
nPrim =      20326 <nPrim> =   406 nPrimMax= 1426 Ptot/Gtot=0.499
nPgoo =      19657 <nPgoo> =   393 nPgooMax= 1394 Pgoo/Ggoo=0.631

It is better but not enough.
Increase to 20 %
Result exactly the same
I keep 10%

===================================================================
Mar27.2014
StvELossTrak rwritten.
Trace of all materials added.
dP/dP0 added and included into derivatives
class StvELossData removed
Short test:
In File(s) 0.log
nFiles =1 nEvents = 50 <nEvents>=50
nGlob =      40524 <nGlob> =   810 nGlobMax= 3233
nGgoo =      31063 <nGgoo> =   621 nGgooMax= 2377 Ggoo/Gtot=0.767
nPrim =      20320 <nPrim> =   406 nPrimMax= 1428 Ptot/Gtot=0.501
nPgoo =      19649 <nPgoo> =   392 nPgooMax= 1399 Pgoo/Ggoo=0.633

===================================================================
Apr1.2014
StvELoss bug workaround. Something wrong with factory
Run full test

In File(s) /star/rcf/prodlog/eval13/log_Stv_Mar_31/trs
nFiles =1094 nEvents = 54091 <nEvents>=49
nGlob =   40578602 <nGlob> =   750 nGlobMax= 3599
nGgoo =   31257173 <nGgoo> =   577 nGgooMax= 2814 Ggoo/Gtot=0.770
nPrim =   20741260 <nPrim> =   383 nPrimMax= 1716 Ptot/Gtot=0.511
nPgoo =   20050690 <nPgoo> =   370 nPgooMax= 1673 Pgoo/Ggoo=0.641

Slightly better than before

Timing test: 3.21

====================================================================
Apr14.2014
vout bug fixed.
But still ELoss from StvELossTrak is different from fEdep. Bigger than 20%
Test Jobs with fixed bug ended. But no big hope that low pt will be correct

In File(s) /star/rcf/prodlog/eval13/log_Stv_Apr_14/trs
nFiles =1094 nEvents = 54303 <nEvents>=49
nGlob =   40732232 <nGlob> =   750 nGlobMax= 3602
nGgoo =   31372770 <nGgoo> =   577 nGgooMax= 2815 Ggoo/Gtot=0.770
nPrim =   20818200 <nPrim> =   383 nPrimMax= 1715 Ptot/Gtot=0.511
nPgoo =   20131590 <nPgoo> =   370 nPgooMax= 1672 Pgoo/Ggoo=0.642