G2p Analysis Minutes

From Hall A Wiki
Revision as of 17:24, 19 March 2014 by Melissac (Talk | contribs) (3/12/2014)

Jump to: navigation, search

Minutes of the weekly analysis meetings



Present: Chao, Jie, Min, Melissa
By Phone: Pengjia, Ellie, Karl

Feature Presentations:

  • Chao:
    • Gave an update on status of optics calibrations. He has finished two more settings. The first setting is 1.1 GeV, 2.5T, transverse, with the "worst"
      septum setting. He included a plot of the simulation result showing the effect of the target field. An event by event simulation was used to determine
      the effective theta and phi angles. There is still a small offset in the phi angle for the central hole; this will have to be considered carefully. The
      uncertainty of the dp calibration is of the same order (10^-4) as previous calibration results that have been shown for other energy settings. He has
      also completed the calibration for the 2.2 GeV, 2.5T, transverse setting, this time with the "best" septum setting. Similar results were seen for this setting,
      though the uncertainty in dp is slightly smaller, as the bending effect is not as large for this setting at it is for the 1.1 GeV setting. Of the 8 total optics
      configurations, 5 have been completed and 1 is in progress. The last two settings should be completed in the next couple of weeks, and then it will take
      ~2 weeks to complete the calibrations for the RHRS. More details can be seen in his slides here.
  • Jie:
    • Gave an update on including radiative corrections in the simulation package. Last time, there was a question about why the un-radiated distribution (from
      a carbon target) was not symmetric. It appears the value for theta used at the event generator to calculate dp is a uniform distribution, which causes the
      strange shape in the distribution. He described two possible methods to get the final distribution; they are similar but the first method is weighted by the
      elastic Born cross section while the second method is weighted by the radiative cross section. Comparing the two methods, the distributions for dE at the sieve
      show a considerable difference. However, if you compare the distributions for the two methods at the event generator, they are almost the same. It's possible
      that the second method accounts for the energy loss dues to ionization and Bremsstrahlung twice. More details can be seen in Jie's slides here.


Present: JP, Jixie, Kalyan, Jie, Min, Chao
By Phone: Melissa, Pengjia

Feature Presentations:

  • Pengjia:
    • Gave an update about the status of BPM calibrations. He is studying the BPM pedestal value for the run 3600~3900. The typical fluctuation of the
      pedestal value is about 1000 channels. However, Pengjia pointed out that the pedestal value of BPM B changed during some runs (like 3714 and 3732),
      especially for the channel 7 of BPM B. JP suggests that multiple pedestal values can be used in the same run to correct those big jump of pedestal values
      and treat the rest small fluctuations as uncertainty. More details can be seen in his slides here.

General Discussion:

  • Jie is working on radiative corrections in the simulation.
  • Chao is working on optics calibration of 1.1GeV and will compare the result with longitudinal optics.


Present: JP, Kalyan, Vince, Jie, Min, Chao, Jixie, Melissa
By Phone: Pengjia

Feature Presentations:

  • Jie:
    • Gave an update on including radiative corrections in simulation. Included a list of definitions for all the terminology and different processes used
      in the calculations. He showed a comparison of using the peak approximation versus the exact calculation for the NH3 target using two different
      radiation lengths. For small values of nu, there isn't a large difference between the methods. He also showed a comparison of using the exact
      calculation method for two different radiation lengths (0.1/0.03), both with and without the multi-photon correction. Without the correction, the ratio
      is pretty flat, but including the correction causes the ratio to drop for smaller values of nu. He included a plot of the total radiative corrections for the
      carbon target (internal and external), and a plot showing the the carbon cross section with and without corrections. It's unclear why the uncorrected
      cross section is not symmetric; it appears to already have some energy loss. Jie will check this and will also include detector resolution and acceptance,
      which aren't included in these plots. More details can be seen in his slides here.
  • Chao:
    • Gave an update on longitudinal optics. He took a closer look at the BPM x fluctuation during the course of a run; it seems the beam position is
      slowly fluctuating over the course of the run. He described the method for matrix calibration; the matrix elements are calibrated by the open angle of
      2 different sieve holes with respect to the scattering point. The fluctuation in the beam position will slightly influence the calibration. The matrix
      calibration can be calculated using geometry relations, so he simulated several different beam positions and calculated the open angle; the results
      showed that the change in angle due to the beam position is small. JP suggested that he should look at the absolute value of the sieve holes, instead
      of the difference between them. He also showed a comparison of the first order matrix for the cases with and without the target field. The theta offset is
      large, which is due to the bend from the target field. JP asked about the effect on the left septum after the right septum broke; according to simulation
      there is ~10% effect. More details can be seen in his slides here.

General Discussion:

  • Pengjia is working on checking the pedestals for all periods for BPM calibrations. The change in pedestals seems to be random. There also seems
    to be an effect where changes in current affect the BPMS. Pengjia says this effect seems to be random, but he will show more details on this next time.


Present: JP, Jixie, Jie, Min, Chao, Kalyan, Melissa
By Phone: Karl, Ellie, Toby, Ryan, Pengjia, Moshe

Feature Presentations:

  • Melissa:
    • Gave a short update on the inconsistency in the yields. There seems to be a correlation between beam position and yield; for p0 = 2.072 GeV/c
      (2.5GeV beam, 2.5T magnetic field) setting, an ~8% drop in yields can be seen when the beam position changes ~3mm in X and ~2mm in Y. She
      also tested the effect of applying different raster cuts, in case the shift in beam caused scraping. The raster cuts did not help to make the yields
      agree. JP suggested that, due to our small scattering angle, a change in beam position can have a significant change in the kinematics. She will
      look more into this, and look at the corresponding effect on the RHRS. Her slides can be seen here.
  • Min:
    • Gave an update on the pointing uncertainty. Previously, she used the carbon ground state and first excited state to determine the uncertainty.
      A landau-gaussian convolution fit was used to fit the two peaks and determine the uncertainty. JP cautioned against using fits in Root, as they
      can be unreliable. Min will take a closer look at this and the determination of the uncertainty. Using this method gives a d-theta of ~0.2mr, which
      is within the requirement. She also showed another method to calculate the uncertainty, using the difference in energy between two nuclei (carbon
      and helium), this time using a Monte Carlo simulation to determine the beam offset. The beam position at the target looks identical for both carbon
      and helium, though JP suggested only using one "blob" (the "blobs" arise from drift in the beam position during the run) in order to get better
      resolution. If only survey results are used, the value for d-theta is ~0.7mr. More details can be seen in Min's slides here.
  • Pengjia:
    • Gave an overview of status of BPM calibrations. The beampackage tools have been updated, he has dumped all the raw BPM data from volatile
      rootfiles to a data pool and have saved all calculated positions to disk. Students can now insert this information into their own rootfiles. He is
      calculating the uncertainty for each run and checking for any weird changes in beam position. His results will be published on the ELOG when he
      is done. A database containing the average position for each run is built along with a script to access the information. He showed global plots
      of the beam position for every run at BPMA, BPMB and the target. He also showed a global plot of pedestal values for the entire run period. Some
      pedestal runs were taken with out any current (real pedestal runs), while some were taken during long beam trips. Currently, he uses the pedestal
      closest in time to each production run as the pedestal value, but JP suggested breaking them up by setting instead. Pengjia is still working on how to
      estimate the uncertainty for the pedestal value. More details can be seen in his slides here.

General Discussion:

  • Moshe has his simulation ready, but needs optics calibrations! JP suggested using straight thru optics results for now.


Present: Jixie, Kalyan, JP, Min, Chao, Jie, Melissa
By Phone: Toby, Ryan, Ellie, Karl, Pengjia

Feature Presentations:

  • Jie:
    • Gave an update on his work to determine the packing fraction using simulation. He is working on updating his simulation to include radiative
      corrections. He is including a correction for the straggling effect, which arises due to "real" bremsstrahlung and ionization. He showed a comparison
      between the peak-approximation method and the exact calculation, for a carbon foil. They agree within 1%, but the difference would probably be
      larger for a (thicker) ammonia target. He also showed the simulated result for the radiative tail in the carbon target, which suggests that the "internal"
      contribution is larger than the "external contribution, except at small values of nu. Finally, he showed the effect of the multi-photon radiation correction,
      which is fairly large at small values of nu. He is still working on optimizing the code. More details can be seen in his slides here.
  • Ryan:
    • Gave an update on the work he's been doing on the radiative corrections for small angle GDH data. At our kinematics (2.1GeV, 6deg), the cross section
      has a strong dependence on the scattering angle, which means the cross section could vary greatly over the entire angular acceptance. The code used to
      produce the elastic radiative tail uses one central scattering angle, so angular acceptance needs to be corrected for. He broke up the acceptance into a grid
      in theta and phi bins, calculated the elastic tail at different angles across the acceptance, and averaged the results. The difference in the cross section using
      the central scattering angle vs. using the average of the reconstructed scattering angles is about 20-25%. He will do a comparison of carbon/nitrogen data
      for 3.7 GeV once he gets the nitrogen data from Vince. He is also writing up a technote documenting this work. More details can be found in his slides here.
  • Chao:
    • Gave an update on the status of optics calibration. A program (g2prec) is written to read in L.gold.* (effective angles) and Lrb.tgt_0_* (bpm) type
      variables as input for projection and drifting functions in the simulation package to calculate the final reconstructed variables. It will be configured by the
      database to deal with different kinematic settings, and will write L.rec.* type variables into the final rootfiles as the final reconstructed scattering angle and
      and momentum. He also showed a diagram depicting the structure of the "do_replay" script, which will be automated to insert all necessary information
      into the final rootfiles. He is currently working on determining the calibration matrix for the 9 different optics settings, which can probably be completed in
      ~1 month, though the uncertainties will take longer. His slides can be found here.


Present: Chao, Jixie, Jie, JP, Min, Melissa
By Phone: Toby, Ryan, Ellie, Karl, Pengjia

Feature Presentations:

  • Melissa:
    • Showed a summary of checks she's done to try to understand the discrepancy in yields for packing fraction runs. Many different tests were
      done testing the cuts (acceptance, PID and raster) as well as checking the multitrack efficiency and comparing the first and second half of each run.
      JP pointed out a few odd things that should be resolved (such as the "spikes" in the reconstructed phi variable, events above the good electrons in
      leadglass 2D plot, etc.) The difference in yields is still not clear, but it's possible that a Moller measurement that was taken between runs 3503 and
      3574 could have altered the beam conditions. She will discuss with Pengjia to see if the Moller measurement had a large effect on beam position.
      More details can be seen in her slides here.
  • Min:
    • Showed an update on her pointing study to determine the central scattering angle. She showed the details of calculating the uncertainty in the
      scattering angle using survey results. She used the assumption that z and x are the same for both theta1 and theta2, which JP warned could in be
      incorrect. Looking at the results for the left and right HRS, the results from pointing and survey agree for the LHRS, but not for the RHRS. The pointing
      uncertainty is different for left and right HRS; this is determined using the difference in the carbon ground state and first excited state, compared to
      the nominal value. Min is currently writing up the details of the pointing study so far. More details can be seen in her slides here.
  • Pengjia:
    • Gave an update on the status of BPM calibrations. He is working on the calibration for the optics run near 3185, but unfortunately there are no straight
      through calibrations with div=2. The closest calibration is from 3/6 with div=3. Looking at the results for run 3185 using this calibration gives a beam
      position of -0.84mm(X) and 2.39mm(Y) at the target. Chao's fitting result for this run was -3.5mm(X). JP asked how much a change in beam position of
      3mm would affect the central angle calibration. Pengjia will try to get the calibration as good as possible, then summarize his work and determine the
      uncertainty for the beam position. Also, he will compile a table of all beam condition changed throughout the run. His slides can be seen here.


Present: Kalyan, JP, Chao, Jie, Min, Jixie, Melissa
By Phone: Toby, Ryan, Ellie

Feature Presentations:

  • Toby:
    • Working on dilution analysis. Gave an explanation of his method to determine the background yield in terms of the dilution runs (carbon,
      empty, dummy). He also needed to create a scaling factor to relate the carbon runs to the nitrogen background. Using Moshe's nitrogen
      simulation he was able to scale the carbon production data (with the "empty" run yield subtracted out) to match the nitrogen yield. He made
      a graphical cut on the carbon and empty dilution to isolate the elastic/quasi-elastic channels so that each channel could be scaled separately.
      He can then use these scaling factors to determine the overall background yield. The current scaling factors result in a background contribution
      that is much too large; the yield is greater than the yield for a production run. This is most likely due to the fact that Toby currently doesn't
      include any radiative corrections, which he will include in his method for next time. The details of his method can be seen in his slides here.
  • Chao:
    • Gave an update on the status of optics calibrations. He is currently working on longitudinal optics for the 2.2 GeV, 5T setting, since Min will
      need this for her pointing study. He first used a simulation to show the effect of the longitudinal field, setting the beam position to the average
      value of the BPM readout. The beam positions actually vary slightly between different optics runs, so he had to approximate the beam position for
      "bunches" of 5000 events. An event by event simulation was used to calculate the effective theta and phi angle. The calibration looks pretty good
      for this setting. The calibration tool kit, which combines the optimizer and the simulation, is almost finalized. He is currently working on updating
      the reconstruction script to calculate the kinematic variables. More details can be seen in Chao's slides here.
  • Jie:
    • Working on updating the simulation package to include radiative corrections. There are two pieces to the correction; the radiation effect and the
      straggling effect. The radiation effect includes the contribution from the dynamical effect (Bremstrahlung), which Jie has approximated as two
      external radiator lengths. Ryan suggested just doing the full internal calculation, as it doesn't require much extra computing power. Since our target
      thickness is ~0.03, the straggling effect is significant. He showed the simulated yields for the elastic nitrogen peak with and without the radiative
      corrections, but neglected to include the radiative tail, which he will include this for next time. He also looked at the simulated yields using the actual
      beam positions for the packing fraction runs 3503, 3574 and 3864. The yields agree within 4.5%, meaning that the differences in beam position is not
      what is causing the discrepancy in the yields of the packing fraction runs. Jie's slides can be seen here.


Present: Kalyan, JP, Chao, Jie, Min, Melissa
By Phone: Toby, Ryan, Karl, Pengjia

Feature Presentations:

  • Melissa:
    • Gave an update on packing fraction analysis. Compared the x/y beam positions for the four packing fraction runs for material 7; one
      run (3727) has unreliable beam position information, but the other 3 runs seem fairly consistent. There is still a large discrepancy in the
      yield for one run (3503). This is the earliest run taken; she will go through the logbook to look for any condition changes that could cause
      this discrepancy. The results for the packing fraction for the other 3 runs seem fairly consistent, but lower than expected. More details can
      be seen here.
  • Pengjia:
    • Gave an overview of his calibration method; the raw signal received in the antenna is compared to the recorded ADC value, and the
      linear region is fit. This can then be rewritten using a diff/sum value and the nonlinearity corrected for. The position is then calibrated
      using data from harp scans. Some of the calibrations were used for runs with different currents, assuming that the beam position
      didn't change for these runs, which JP suggested was a dangerous assumption to make. There were several periods where the pedestal
      was different between the calibration and production runs. This fluctuating pedestal could change the resulting beam position by ~2mm.
      For next time, Pengjia will summarize the assumptions made for these calibrations so the uncertainty can be assessed for each one. More
      details of his calibration method can be seen in his slides here.
  • Min:
    • Gave an update on the pointing study to determine the central scattering angle. She showed the equation for elastic scattering
      (which includes energy loss); she will look at the difference in the scattered energy for two different nuclei. She included two different
      methods for calculating the pointing uncertainty. The first method uses the carbon-12 ground state and first excited state, and gave
      a result for d-theta/theta <1%. JP suggested Min will probably need to combine left and right data to get the precision we need. The
      second method combines various uncertainties (from survey, beam_x, etc.) and uses 2 different scattering angles; one including the
      z-offset and one without. Using just one scattering angle to compute d-theta gives an uncertainty of ~2mr, using the quotient of the
      two scattering angles gives an uncertainty of ~0.2mr, and using the difference of the two angles gives an uncertainty of ~0.04mr. She
      still needs to include input for the incoming beam angle, and include the uncertainty from energy loss. More details can be seen in her
      slides here.

General Discussion:

  • Ryan:
    • Working through Rosetail to understand radiative correction calculations/code.
  • Toby:
    • Looking at scaling carbon data to nitrogen for dilution analysis.
  • Chao:
    • Working on longitudinal optics calibration.


Present: Kalyan, Chao, Jie, Min, JP, Jixie, Melissa
By Phone: Ellie, Toby, Ryan, Karl, Alexandre, Pengjia

Feature Presentations:

  • Jie:
    • Gave an update on using the simulation package to determine the packing fraction. He showed multiple methods for extracting
      the packing fraction, which account for different contributions to the overall yield. He used the g2psim package to get relative cross
      sections for each contributing material. Next he will work on including the radiative corrections to the elastic peak and include the
      calibrated beam position information (once it is ready). Details for the methods used can be seen is his slides here.


Present: Kalyan, Chao, Jie, Min, Melissa
By Phone: Ellie, Toby, Ryan, Karl, Alexandre, Pengjia

Feature Presentations:

  • Melissa:
    • Gave a short update on yields for packing fraction runs. For one run (3864), the ungated livetime is significantly different then
      the helicity-gated livetime. Using the average of the helicity gated LT values for the normalization makes this run agree better with
      other packing fraction runs (3574 and 3727). She will check the beam position for these runs to see if this could have a large effect
      on the yields. Her slides can be seen here.
  • Chao
    • Gave an update on the simulation package. Previously, the real theta and phi were not determined event by event, but for an entire
      run. Since we know the position of the reaction point and the sieve hole, and the momentum of the electron, it is possible to use a
      bisection method to determine the real theta and phi for each event. With this update it will be easier to determine the calibration for
      different dp settings. Next he will work on finishing the optics calibration for the first setting with the target field. His slides can be
      seen here.
  • Min
    • Gave an update on pointing analysis. The calculation was done using the March 14th calibration, using a carbon foil in liquid helium.
      Using the carbon-12 ground state and first excited state to do the calibration of d(delta-E'), the results for the LHRS (4.40MeV) and
      RHRS (4.42 MeV) agreed within the uncertainty with the nominal value (4.44MeV). However, including other uncertainties (survey and
      beam x), add up to ~0.11 deg uncertainty in the scattering angle. She concluded that using the carbon foil in LHe is not ideal for a
      pointing study. Next she will try CH2 run in longitudinal target field setting, once the optics calibration is complete for this setting. Her
      slides can be seen here.
  • Pengjia
    • Gave an update on the status of the BPM calibrations. He has finalized the calibrations for data taken from April 11th to May 18th,
      which includes three different BPM gain settings. For settings before April 11th, the process is more complicated (larger pedestals,
      auto-gain), but he will try to finalize the calibrations soon. He will provide a technote of his procedures and results. More details can
      be seen here.

General Discussion:

  • If any students are interested in giving a talk at the APS April Meeting, please circulate an abstract to the group. The deadline for abstract submission is this Friday, January 10th.

June-Dec 2013


Jan-May 2013


April-Dec 2012


Jan-March 2012


July-Dec 2011


Jan-June 2011