DRAGON analyzer

summary

This analyzer has been tailored for the DRAGON experiment. Most of this program was originally written by Jonty Pearson. The latest enhancements support making graphical cut on one histogram to determine if related histogram is filled.

The syntax for the analyzer input file is similar to carpere7's input file. If you are familiar with those files, then you'll have a significant advantage in understanding how to make cuts in the analyzer using these enhancements.

requirements

The dragon analyser makes use of the following:

A mysql server with a database named 'Online' containing detailed information mapping ADC and TDC channels to each detector [REQUIRED].
A MIDAS data file (with the 'mid' extension) [REQUIRED].
A MIDAS odb file containing the configuration parameters for analysing the midas data file [usually REQUIRED]. The ODB field /Analyzer/ConfigFile determines which input file is parsed to make cuts (see next item).
A text file defining which cuts to be used and in what manner [OPTIONAL].
One or more text files, each containing a root script for a graphical cut [OPTIONAL].

command line

If you confuse the MIDAS analyzer enough you'll get the following:

valid options are:
  -c <filename1>   Configuration file name(s). May contain a '%05d' to be
     <filename2>   replaced by the run number. Up to ten files can be
         ...       specified in one "-c" statement
  -d               Debug flag when started the analyzer fron a debugger.
                   Prevents the system to kill the analyzer when the
                   debugger stops at a breakpoint
  -D               Start analyzer as a daemon in the background (UNIX only).
  -e <experiment>  MIDAS experiment to connect to
  -f               Filter mode. Write original events to output file
                   only if analyzer accepts them (doesn't return ANA_SKIP).

  -h <ostname>>    MIDAS host to connect to when running the analyzer online
  -i <filename1>   Input file name. May contain a '%05d' to be replaced by
     <filename2>   the run number. Up to ten input files can be specified
         ...       in one "-i" statement
  -l               If set, don't load histos from last histo file when
                   running online.
  -L               HBOOK LREC size. Default is 8190.
  -n <count>       Analyze only "count" events.
     <first> <last>
                   Analyze only events from "first" to "last".
     <first> <last> <n>
                   Analyze every n-th event from "first" to "last".
  -o <filename>    Output file name. Extension may be .mid (MIDAS binary),
                   .asc (ASCII) or .rz (HBOOK). If the name contains a '%05d',
                   one output file is generated for each run. Use "OFLN" as
                   output file name to creaate a HBOOK shared memory instead
                   of a file.
  -p <param=value> Set individual parameters to a specific value.
                   Overrides any setting in configuration files
  -P <ODB tree>    Protect an ODB subtree from being overwritten
                   with the online data when ODB gets loaded from .mid file
  -q               Quiet flag. If set, don't display run progress in
                   offline mode.
  -r <range>       Range of run numbers to analyzer like "-r 120 125"
                   to analyze runs 120 to 125 (inclusive). The "-r"
                   flag must be used with a '%05d' in the input file name.
  -s <port>        Start ROOT histo server under <port>. If port==0, don't start server.
  -R               Start ROOT interpreter after analysis has finished.
  -v               Verbose output.
  -w               Produce row-wise N-tuples in outpur .rz file. By
                   default, column-wise N-tuples are used.

In this case, one of the most useful options is the ability to choose which ODB file (what MIDAS calls a 'Configuration file') to load when reanalyzing a MIDAS file. Since the analyzer uses the ODB to know which input file to use when determining cuts, this allows you to change which input file to use when reanalyzing a run.

Example:

~/src/E810/bin/analyzer -c cuts/Mg24_234AkeV_6plus_7.7T.odb -i ../E810/run18320.mid -o media/test.root

example

Test these files by downloading them into a 'cuts' directory and using the command:
analyzer -c cuts/his18320.odb -i cuts/his18320.mid -o analyzerDemo.root

supported histograms

All of the following histograms are supported for analyzer based cuts. Cuts can be made on singles or coincidences. This list also defines the list of histograms which can be populated upon passing a cut.

tag	description
BgoE	Energy vs. Bgo number
BgoE0	Largest detected energy signal (per event) vs. Bgo number
BgoE0E1	E0 vs. E1 (second largest energy signal)
BgoPosition	A 'bear skin' plot of Bgo hit position
GrTofHiTof	BGO array TOF vs. Heavy Ion TOF
IcE	Energy vs. IC anode
IcEdE	IC anode 0 vs. anode 1
IcEtotE0	total IC energy vs anode 0
IcEtotE1	total IC energy vs anode 1
IcEtotE2	total IC energy vs anode 2
IcEtotE3	total IC energy vs anode 3
IcEtotE4	total IC energy vs anode 4
IcSum	total IC energy
IcSumVsTAC	total IC energy vs. local MCP TOF
McpTOF	local MCP TOF
McpXY	spatial position of MCP hit

input files

The input file is broken into blocks, each representing a single command, contained within open and close tags. See below for a list of blocks supported.

open tag	description
cut	define a graphical cut on a histogram
group	define a group of histograms, requiring the dragon events to pass one or more cuts
version	alters processing of the input file based on a version

And here is a more complete description of the syntax for each block:

cut

Define a graphical cut on a histogram type.

tag	parameters	description
cut		open tag
name	<name>	referenced by 'group' tags
object	<name>	name of the cut within the root script
script	<file name>	file name containing the root script
type	<histogram type>	a single histogram type taken from the list above
end		closing tag

Example:

cut	
	script	cuts/Mg24_234AkeV_6plus_7.7T_cut1.C
	object	cut1_77
	name	cut1
	type	IcSumVsTAC
end

group

Define a histogram group -- each event must pass the cuts on the group to be included in the histograms.

tag	parameters	description
group		open tag
name	<prefix>	All histograms have this prefix
data	<type>	Supported types are 'single' or 'coincidence'
folder	<name>	All histograms are placed within this folder (under DragonEvent)
cuts		A block of cut names (terminate with 'end')
histograms		A block of histogram types taken from the list above (terminate with 'end')
end		closing tag

Example:

group
	name	hsCut1
	data	single
	folder	Cut1_Singles

	cuts
		cut1
	end

	histograms	
		IcE
		IcEdE
		IcEtotE0
		IcEtotE1
		IcEtotE2
		IcSumVsTAC
	end
end

Example:

group
	name	hcCut1
	data	coincidence
	folder	Cut1_Coincidence

	cuts
		cut1
	end

	histograms	
		IcE
		IcEdE
		IcEtotE0
		IcEtotE1
		IcEtotE2
		IcSumVsTAC
		BgoE
		BgoE0
		BgoE0E1
		BgoPosition
	end
end

version

The version tag (not really a block in this case) is supported in case future input files have changed their syntax beyond the point of backwards compatibility. Presently, only version 1 is supported.

Example:

version 1

LE
13 June, 2008