1) login into PDSF 2) say you are at directory /auto/u/steph 3) cp -r /auto/romero/test t 4) ls -l t total 24 drwxr-xr-x 2 romero rhstar 4096 Jul 9 12:13 fits -rw-r--r-- 1 romero rhstar 8922 Jul 9 12:28 y0template.lsf -rw-r--r-- 1 romero rhstar 4483 Jul 9 12:11 yall.scr 5) Directory "t" (or whatever) is the directory you want to do the fits from. Initially, directory t/fits is empty: ls -l t/fits total 0 6) pico file y0template.lsf and replace "test" by "t" (or whatever the directory that contains directory "fits" is called) in lines 11 and 15. 7) To test that you can get fits, without modifying file y0template.lsf, do ... source yall.scr r 0 0 -1 8) The directory t/fits/ny0av6r is created. In about 5 minutes it contains quite a few files. The important files are, for pions, fitspiny0av6r.root -- this contains all dedx histos and normYields (as in .ps) fitspiny0av6r.ps -- .ps file, as in the past. Same info as above, for printing. ny0av6r/piny0av6r.log -- log file, lots of information on fitting and assumptions ny0av6r/piny0av6r.sum -- summary text file (from the .log file above) showing cuts used, Nevents, chi2/df, etc. At the end is normYield for pi To check that things worked, except for the dates, the file piny0av6r.sum should be identical to the one I ran: diff fits/ny0av6r/piny0av6r.sum ../mini2/fits/ny0av6r/piny0av6r.sum < *** Start at Date : 9-Jul-2001 Time : 12:26:37 *** --- > *** Start at Date : 8-Jul-2001 Time : 16:49:19 *** 9) After a succesful test, you can run your own fits on the different histograms available as described in /auto/romero/mini2/directories.txt To do that, just change any the parameters a0-a23 in y0template.lsf. For fits only, as indicated here, the actual files used are picked from the directories as described in the directories.txt file. 10) For convenience, I include here the directories.txt file [pdsflx001] ~/mini2 > cat directories.txt Table 1.- Files at PDSF available for fits. centra- Files z-vtx mPoints> Nevents* dEdx key directory y-range lity cut cut scale presently (y=0,charge<0) keV/cm** Jul 09,01 ----------------------------------------------------------------------------- A cent none 25 17346 24 t ny0av24t -1.0 to 1.0 A cent none 25 17346 6 c ny0av6c -1.0 to 1.0 A cent 011000 11000 is the minimum value for mCtbEvt used in the trigger. See Trieu Mai's plots in his web page. In the run I used mCtbEvt>8500, but this obviously doesn't do anything. * Nevents is defined as the number of events surviving the cuts A&&V, where A= mHeader.mZdc<66&&mHeader.mCtbEvt>11000 V=mHeader.mVertexPosition.mX3>0&&mHeader.mVertexPosition.mX3<50) (for the vertex cut between 0 and 50). If one changes the minimum mPoints for each track, the number of events doesn't change. ** for scale 6 keV/cm, number of bins in dEdx is 360 (linear plot in dEdx) for scale 24 keV/cm, number of bins in dEdx is 1440 (log plot in dEdx)