Splitting mass in small cells - after debugging

- Akio's study: www.star.bnl.gov/protected/spin/akio/fms2015/showershape.html
- A bug was found which created in recent shower shape update:

I put those values as default (Zhanwen's shape, and 0.8/0.6).
By default, without calling setScaleShowerShape() nor setShowerShapeWithAngle(), one should get the parameters right.

This also fixes a bug where it always uses large cell shower shape parameters with Zhanwen's shower shape and
when setMergeSmallToLarge option on. So effectively Zhanwen's shower shape for small cell was at ~x1.3 wider than intended,
and about ~x2 too wider than this study tells me to put (and only for small cells, and only when setMergeSmallToLarge is on).
This was causing massive pulls of photons towards center.

- Akio, Apr. 16, 2019


- Tested same QA codes after running a Scheduler with debugged libraries (StEvent, StFmsHitMaker, StFmsPointMaker, and StFmsUtil/)
- For the better comparison same type of figures "before" debugging also will be shown
- Only detId 10 (Small cell north) will be shown: detId 11 is NOT very different

  1. Mass vs. pair E, separated by Zgg in stepwidth of 0.1

   1-a. Before debug (reported before)

   1-b. After debug

 2. Mass vs. pair E, incremental Zgg in stepwidth of 0.1

  2-1. Before debug

   2-2. After debug

 3. Distance between points, separated by Zgg in stepwidth of 0.1

  3-1. Before debug

  3-2. After debug


- Tested points' XY distribution if those "grid" like structure still exists
- Weak concentration still exists, but much eased. I think it's reasonable enough

 4. Points' XY distribution

  4-a-1. All, before debug

  4-a-2. All, after debug

  4-b-1. Large cells, before debug

  4-b-2. Large cells, after debug

  4-c-1. Small cells, before debug

  4-c-2. Small cells, after debug


 5. Mass vs. Zgg, separated by pair E in stepwidth of 10

  5-1. Before debug

  5-2. After debug

  * 6.  Mass vs. Zgg, separated by pair E in stepwidth of 10, for detId = 8

  6-1. Before debug

  6-2. After debug