H2D HITFINDER SWITCHES

Here are the h2d switches as of version 1.20 (released 4 Dec 1997)
Also listed are the default values for Au+Au
and the version in which the switch was introduced

For versioning information, see here
Debugging information switches (does not affect performance)
Switch name value type Au+Au pass1 default value Introduced in version numberComment
dbg_histo
	I*4
	0
	?
	1 --> output 2d box histograms of ADC vs
				time and pad for each padrow
dbg_printf
	I*4
	1
	1.19
	nonzero setting gives various non-error
				messages from h2d - useful for debugging
dbg_rms
	I*4
	0
	1.0
	outputs RMS in pad and time for isolated
				peaks - designed to study diffusion
dbg_tune_triage
	I*4
	0
	1.0
	makes a "training file" for triage cut - 
				contains RMS values for overlapping and
				non-overlapping clusters
dbg_visualize
	I*4
	0
	1.0 (?)
	for a given cluster of peaks,
				creates 3D plot of data (ADC-pad-bucket)
				and also what the current fitting model
				of that cluster looks like, for comparison.
				Interactively steps through 2d fitting.

General data correction/cleanup (pertinent to both passes)
Switch name value type Au+Au pass1 default value Introduced in version numberComment
nbuck_cleanup
	I*4
	3
	1.14
	number of buckets from bottom of
				TPC on which to perform pixel cleanup
(see here for discussion.)
cleanup_maxocc
	R*4
	0.25
	1.14
	maximum fractional occupancy allowed on
				lower nbuck_cleanup buckets in TPC
(see here for discussion.)
symmetrize_row51
	I*4
	1 (2,4 GeV)
0 (6,8 GeV)
	1.14
	1=enable, 0=disable, event-by-event
				 cleanup of the noisy half of padrow 51
(see here for discussion.)
satval
	I*4
	2
	1.17
	this is what hits.satf is set to when
				an ADC hits 4096 in a hit.  We set it to
				2 because JTF is hardwired to handle the
				satf=1 hits seperately, which I did not like.

Switches relevant to the first pass
Switch name value type Au+Au pass1 default value Introduced in version numberComment
triage_cut
	I*4
	0
	1.0
	nonzero -> perform RMS cut on clusters
				before attempting deconvolution.  Never
				implemented :(
signif_cut
	I*4
	1
	1.13(?)
	0 -> perform peak/valley cut in determining
				significance of peak.
1 -> perform peak-valley cut in determining significance of peak.
onepass_only
	I*4
	0
	1.10(?)
	nonzero -> do not store pixel information
				for use in second pass (time/memory saver)
breakup_cut
	I*4
	4
	1.0
	maximum number of peaks allowed in cluster
				before cluster breakup attempted
thresh_init
	I*4
	10
	1.0
	initial value of ADC threshold in cluster
				deconvolution/breakup
thresh_incr
	I*4
	5
	1.0
	amount ADC threshold is raised in each cluster
				breakup step
ntry_breakup
	I*4
	10
	1.13
	number of attempts to breakup clusters
peakval_cut
	R*4
	10
	1.0
	peak/valley or peak-valley cut, depending on
				value of signif_cut

Switches relevant to the second pass
Switch name value type Au+Au pass1 default value Introduced in version numberComment
remove_hits
	I*4
	0
	1.0(?)
	delete hits that were untracked (from
				first tracking pass) from HITS table
decon_small
	I*4
	1
	1.0(?)
	1 -> do deconvolution for clusters w/#peaks
				<max_peaks_fit 
nofit_isolated
	I*4
	1
	1.0(?)
	1 -> do not fit isolated peaks to extract area -- just add up ADC values
onetrk_pass
	I*4
	0
	1.0(?)
	1 ->
			do not zero out HITS.TRACKID - use if no 2nd tracking
max_peaks_fit
	I*4
	4
	1.0(?)
	 max #peaks in cluster that we try to fit.
If a cluster has more peaks than this, we do not get dEdx from those peaks (hits)
keep_yave
	I*4
	1
	1.20
	force average HITS.Y to be same as first pass (and 1dh)

For information about modes, see here.
Mode Switches
Switch name value type Au+Au pass1 default value Introduced in version numberComment
mode1
	I*4
	0
	1.0
	First pass mode
mode2
	I*4
	6
	1.0
	Second pass mode
internal_mode
	I*4
	-101
	1.0
	Used internally by hitfinder-- do not attempt to modify!!!

lisa@mps.ohio-state.edu