Hello,
in preparation of our meeting (tomorrow) I have a few announcements
_ The data files are now transferred to the spuds every 3 minutes
_ NOTHING has changed in commis/phase1 which is still the actual analysis
directory
However, I take the job of making sure that the online analysis is
homogenous very seriously. So I have a number of points to discuss with
you. In short I propose a commis/phase2.
It alread exists in a tmp version and we should discuss when to switch
(come to the meeting) when we switch (hopefully soon), we'll stick to it.
I think that if we agree on what macros we need and how to maintain them,
it will be possible to make an analysis man page for people on shift
**********************************************************************
in commis/phase2
_ ntuple.C is upgraded: all ntuple data is pedestal subratcted. That
does absolutely not mean that you can not look at pedestals (see
below). However we need to converge on a common set of cuts, not 64
adc cuts. Also, offsets make presentation of data problematic.
_ ntuple.C is automatically run and an ntp and eep ntuples generated.
_ ntuple.C contains branches, which slightly affect the syntax of
your commands. Else we'll need to change code when the NC
come on line, in the middle of running. I think it is
better to do all testing in phase2 before hand.
_ All parameter files stay only in Blast_Param (including e-map)
and are archived the usual way
_ with that, and absolute pathnames, we can organize the analyze
directory in subdirectories. That is easier than carrying over links
(I think of CVS). Of course part of the path can be host depedent
and be set in an environment variable.
_ names of raw data macros start are prepended by "raw_".
Note I've carried over
raw_adc.C : plots adcs, cut on TDC
raw_tdc.C : plots tdc
raw_pedestals.C: plots pedestals
retime/raw_retime_startc_in?16.C: the retime analysis macro
dump.C
_ for analysis of the general ntp ntuple we have
tof_eff.C
cc_eff.C
"offset subtracted" show_tdc.C, show_adc.C
_ for analysis of data from the reduced eep ntuple we have
show_tof.C (used to be show_ptof.C, shows L-R, R-L time)
show_coplanarity.C (definitely needed, but a 10 min job)
adc_adc.C
ep_counts.C
_ all utilities macros (util.C, CleanUp.C, CodaRead.C ...) go to a subdir
Common rules:
_ if you need to change a cut you should *not* make another macro
Edit the file. It's good to read what it is in it
_ All macros should -as much as possible- stick to having a init.C
command line interface, a variable run number and so on..
_ Likewise for the run number and the date
_ All plots, .ps, .eps *reside* in the results directory. They'll be
zapped in commis/phase1. BTW, for some macros we may generate a .ps
automatically (?)
_ there is a devel directory. That should be used for development, which
you may do on shiff. Only final results go back to the main directory
-- ________________________________________________________________________________ Tancredi Botto, phone: +1-617-253-9204 mobile: +1-978-490-4124 research scientist MIT/Bates, 21 Manning Av Middleton MA, 01949 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST