Dear all this is a letter that I sent you before my departure for my vacations. Because I am not sure that you received it (I sent it from and to my jlab.org account but it seems that I still did not receive it on my jlab computers), I put all the documentations in this folder hoping that someone can read it All files quoted in this letter are in the same folder Best regards Guido From urciuoli@jlabs1.jlab.org Sat Aug 4 08:57:16 2007 Date: Sat, 4 Aug 2007 07:32:09 -0400 (EDT) From: Guido Urciuoli To: kaniol@calstatela.edu, saha@jlab.org, jose@nuc2.fis.ucm.es, cornejo@jlab.org, joaquin@nuc1.fis.ucm.es, camsonne@jlab.org Cc: urciuo@jlab.org, urciuoli@ax0rm1.roma1.infn.it Subject: optical data base and raster corrections Dear all first of all I apologize for my delay. Unfortunately ifarml3 did not work until a couple of days ago and even my office computer in Rome started misfunctioning. As a matter of fact I send mow this email from my jlab address. Please answer my email both to my jlab addrees ("urciuoli@jlab.org") and to my Rome address ("urciuoli@ax0rm1.roma1.infn.it"). Because of ifarml3 misfunctioning I was able to perform my checks only recently and to send this email only today. I apologize again. I do not know if Joaquin and/or Juan have time or will to check what is written on this email in this weekend. Unfortunately I am leaving for vacation next monday. I hope (but not sure) to be able to read emails during my vacations. I will be back in Rome on August 24th (from that day on I will be certainly able to read and answer emails). This email is pretty long. I hence divided it in four nearly indipendent sections: "OPTICAL DATA BASE" "RASTER CORRECTIONS" "RESULTS" "CONCLUSIONS" Everybody can read what is most interested on -------------------- O P T I C A L D A T A B A S E ------------ a new optical data base is now in the directory "/w/work5309/e06007/guido/DB/20080727" or equivalently in the directory "/work/halla/e06007/disk1/guido/DB/20080727" of ifarml3. Using this data base you should be able to obtain the results I showed you during our last phone meetings or the results described below in the section "RSULTS" of this letter. NOTE: the data base works well inside the scattering variable limits: -0.045 < dp < 0.045 - 0.035 < phi < 0.035 - 0.05 < theta < 0.05 (for both arms) I will work after my vacation to extend these limits. ------------ R A S T E R C O R R E C T I O N S ------------------ I recognized that unfortunaltely, in our case, for a good resolution, is not possible to use simple first order formulas of the variable "ReactPt_L_y" (this variable basically describes the position on the target of the incident beam). This forbids us to use standard ANALYZER raster corrections. As a matter of fact, basically, (as far as I know) ANALYER at the moment just allows to change the coefficients in the first order expressions of the variable ReactPt_L.y that it uses to change the values of the scattering variables "DP" and "THETA". The only way for us to have reasonable good raster corrections is hence to plot the ANALYZER missing energy variable ("Skxceb.emiss") with the addition of a polynomial expression of "ReactPt_L_y" that corrects for the beam position on the target. This method is conceptually completely equivalent to the polynomial expressions of scattering variables that we add to the missing energy during our data taking shifts to improve the resolution when plotting the missing energy itself. Those polynomial expressions were 100% equivalent to a standard optic data base in focal plane variables and that optic data base (further improved) is now in the directory "/w/work5309/e06007/guido/DB/20080727" as explained in the section "OPTICAL DATA BASE" above. In the same way, the polynomial expression of "ReactPt_L_y" variable to be added to the missing energy would be 100% equivalent to a "data base" that would take in account higher order raster corrections. As explained above, higher order raster corrections are not implemented in ANALYZER so far The polynomial expresion to be added to missing energy to perform "reasonable good" raster corrections is: Polynomial = -1294*(ReactPt_L.y) - 19220*ReactPt_L.y**2 + + 3800000.*ReactPt_L.y**3 + 190000000.*ReactPt_L.y**4 in other word instead of plotting the ANALYZER variable "SKxceb.emiss*1000" to have the missing energy in MeV, one has to plot the expression: "SKxceb.emiss*1000 - 1294*ReactPt_L.y - 19220*ReactPt_L.y**2 + 3800000.*ReactPt_L.y**3 + 190000000.*ReactPt_L.y**4" instead. Two more words on this polynomial expression. Unfortunately it was determined using run 1195. This run has the best kinematical setting and target (Carbon) for optic study, but was performed in a moment when raster currents were still to be fixed (see below). In addition Bodo's routine "get_rast_const.C", that gives raster constants on a sinle run basis, seems not work any more (incompatatibility with the new ANALYZER/ROOT version? I have to check). That means that the polynomial epression given above, that work reasonably for run 1195 and with raster constants equal to those contained in the directory /w/work5309/e06007/guido/DB/20080727" at the moment, could be less effective when analyzing other runs and/or with orther raster constants. I will check this after my vacation. My experience during data taking shifts is that raster correction formulas (not numeric raster correction of course!) do not depend so much on these factors. As a matter of fact we were able to use a rastter correction formula determined in very poor conditions for all the experiment. Nethertheless ... you never know. Of course you are free to improve (or even completely change) the polynomial expression if you see that in this way raster corrections can be improved. ---------------------- R E S U L T S ---------------------------- as explained in the section "Raster Corrections" above the results refer to the run 1195 (Raster ON, Crabon Target) The raster currents at that time were not good (see again "Raster Corrections" section about the reason I decided to analyze this run all the same) This is shown by the two first files attacched herewith "run_1195_ReactPt_L_x.ps" and "run_1195_ReactPt_L_y.ps" that show the plot of x and y beam position respectively. (I think you remember very well these plots that show the raster current misfunctioning at that time). a) the missing energy spectra without raster corrections: (file "run_1195_0.ps") b) the missing energy spectra with the polynomial expression described in "RasterCorrections" section of this email added to perform raster corrections: file "run_1195_1.ps" c) same as b) but with RectPt_L.x < 0.009: file ("run_1195_2.ps") d) same as b) but with abs(RectPt_L.y) < 0.0015: file ("run_1195_3.ps") e) same as b) and c) but with abs(RectPt_L.y) < 0.001: file ("run_1195_4.ps") The files show that 1) because of the bad raster currents, the resolution could not be good if you do not apply cuts on beam incident position. This can be seen just looking at the "run_1195_1.ps" and "run_1195_2.ps" plots. "run_1195_2.ps" is obtained from "run_1195_1.ps" just cutting awaay events with extremely high x beam position. Un unphysical peak at the left of the main peak present in "run_1195_1.ps" disappears in "run_1195_2.ps" and the resolution in "run_1195_2.ps" is evidently better. The events cut away in the plot "run_1195_2.ps" ARE UNRECOVERABLE (you cannot correct for the variable ReactPt_L.x) 2) the polynomial expression does not work well for all the range of ReactPt_L.y in the run 1195. This can be seen from the plots "run_1195_3.ps" and "run_1195_4.ps" obtaining with a decreasing range of this variable. Nethertheless the result obtained in "run_1195_4.ps" (abiut 900/950 keV of resolution) is encouraging NOTE: all plots have a 100 keV binning and were performed with the following cuts in both HRS arm scattering vaariables -0.045 < dp < 0.045 - 0.035 < phi < 0.035 - 0.05 < theta < 0.05 ) ----------------------- C O N C L U S IO N S ------------------------- Satisfactory results were achieved when analyzing the run 1195 with a new optic data base and with new raster corrections, providing the range of the variable "ReactPt_L_y" (y beam position) is not too big. The raster corrections have still to be checked (and probably improved) testing other runs performed when raster currents were fixed. In any case we can hope to have 1 MeV (or maybe something less) resolution for all the runs. The last two files attache herewith "Cut.macro" and "CarbonPlot.C" allow to obtain the reultsdescribed in rhe section "Results" Just type in ANALYZER environment (after loading the root file obtained with the new data base) the three commands: ".x Cut.macro" ".L CarbonPlot.C" "backsig()" Best regards Guido [ Part 2, "" Application/POSTSCRIPT 14KB. ] [ Unable to print this part. ] [ Part 3, "" Application/POSTSCRIPT 14KB. ] [ Unable to print this part. ] [ Part 4, "" Application/POSTSCRIPT 21KB. ] [ Unable to print this part. ] [ Part 5, "" Application/POSTSCRIPT 20KB. ] [ Unable to print this part. ] [ Part 6, "" Application/POSTSCRIPT 19KB. ] [ Unable to print this part. ] [ Part 7, "" Application/POSTSCRIPT 18KB. ] [ Unable to print this part. ] [ Part 8, "" Application/POSTSCRIPT 19KB. ] [ Unable to print this part. ] [ Part 9, "" Text/PLAIN (Name: "Cut.macro") 178 lines. ] [ Unable to print this part. ] [ Part 10, "" Text/PLAIN (Name: "CarbonPlot.C") 11 lines. ] [ Unable to print this part. ]