####################################################################
#
# ILDPerformance:
#
# collection of scripts and steering files for creating 
# canonical ILD performance plots
#
# authors: Yorgos Voutsinas, Alonso Guerrero Llorente
#
###################################################################


BUILD instructions:

mkdir build; cd build
cmake -C $ILCSOFT/ILCSoft.cmake .. ; make install


1) Resolution plots

Run the runAll.sh

./tracking/scripts/runAll.sh

This scripts simulates single muon tracks, at 4 different polar angles and 8 different momenta
using Mokka particle gun. Then it reconstructs them using the std reconstruction and then,
runs a Diagnostics processor which writes down information to a root tree.
The produced 32 root files are stored at tracking/Results/Analysis.
Finally runs a root macro, which initiates a ROOT GUI where the user can plot IP and momentum
resolution.


If one wants to modify the detector configuration, he has to edit the mokka section of the runAll.sh


The user can reproduce the resolution plots by running
the corresponding macros located at tracking/macros.


By default is the full tracks (MarlinTrkTrack collection) which are studied.
To change to another track collection, the Diagnostics.xml steerin file has to be edited.
In particular, the parameter TrackCollection of the RecoMCTruthLinker processor and the parameter
StudiedTracks of the Diagnostics processor need to be changed.
These two parameters need to be consistent.


It is not needed to run the whole simulation-reconstruction chain again in order to examine a different track collection.
One can simply run the runAnalysis.sh scripts, which reads the already reconstructed lcio files
(provided of course that the relevant track collection has been econstructed)


One can separately run the simulation, reconstruction and analysis steps by using the scripts that are
loated in directory tracking/scripts.




2) Tracking pattern recognition

The Diagnostics processor can be used in order to study the efficiency and fake hit rate of a tracking algorithm.
Script PatRec.sh runs the Diagnostics processor over a reconstructed lcio file. 
The user has to provide the name and path of the reconstructed file by editing the PatRec.sh and modifying the
INFILE variable.

To run analysis on a physics sample, the processor parameter PhysSampleOn should be set to true.

Again, are the full tracks the ones that are studied by default. To change that, the user needs to follow same the instructions
given above.


The track is considered as found if >= 75% of the track hits belong to the dominant MC particle. If one wants to focus on Silicon tracking, one 
needs to set the processor parameter SiTrkEffOn as true. Then the track is considered as found if >= 75% of the track hits belong to the dominant MC 
particle && the track has at least 4 silicon (VXD - SIT) hits. The user can change the definiton of found track, as well as the composition of the 
studied MCParticle subsample by modifying the corresponding processor parameters in the steering file Diagnostics.xml


The PatRec.sh will initialise a root session where with the help of a GUI the user can plot what he wants.
All plots and relevant info is stored in the root file.




3) Track fitting


The PatRec.sh script can provide useful information for track fitting as well.
Assigning a single tracks file to the INFILE variable, the user can plot the track parameter
residuals and pulls.



4) Track - hit residuals


On going...