Mrenna Matched Datasets |
|||||||||
Introduction (can be skipped): The datasets are built from samples of Madgraph events for W+0,1,2,3, and 4 jets and Gamma*/Z+0,1,2 and 3 jets. These exclusive samples (i.e. describing a fixed number of partons) are combined into inclusive samples (i.e. describing any number of partons/hadrons) using Pythia. The method for doing this is not the same as the method of Catani et al used for e+ e- collisions, (know as CKKW), but is "equivalent." The final product is provided to you, the user, as fully hadronized events in the MCFIO format, residing on the Patriot (Physics Analysis Tools Required to Investigate Our Theories) enstore, and described with metadata in SAM. |
|||||||||
What is the CKKW method (can be skipped)? In short, the CKKW "idea"
is to map a W+n parton event into a parton shower history.
All we have is the kinematics and |
|||||||||
How are the matched samples different from CKKW (can be skipped, read above for background)?
|
|||||||||
Where are the samples in Enstore space (can be skipped if using SAM)?
|
|||||||||
Where are the samples in SAM space?
|
|||||||||
Example Metadata Listing:
"sam get metadata
--file=1.Z0run1.1.io"
|
|||||||||
What are the datasets? There are 4 datasets of interest (for today).
Why are there only 3 hard jets for Gamma*/Z? This was a practical issue resulting from the fact that I allowed Gamma* with a mass cut M_ll > 20 GeV. This could be improved in the future, particularly if one is only interesting in M_ll ~ M_Z. |
|||||||||
What are the individual files within the datasets? Lets concentrate first on dataset 1 noted above.
|
|||||||||
How should one weight events? The table below gives the crucial information. Consider the case where you take 50k events from 1.Z0run1.1.io and 1.Z0run1.2.io. By looking at the table, you will see that these should correspond to X=262.6 pb, or a weight of X/50k per event. You now add this to 75k events from 1.Z1run1.1.io, 1.Z1run1.2.io, and 1.Z1run1.3.io, corresponding to Y=94.0 pb. Each event then has the weight Y/75k. If you now made a histogram from both sets of events using this weight, the total area of the histogram should be X+Y=356.6 pb, and there should be 125k entries. Now, continue to increasing multiplicity (Z2runX, Z3runX) until you have used up all the cross section. If you do not want to deal with weighted events, you can choose the event sample that corresponds to the smallest integrated luminosity, use ALL those events, and then use N = sigma * Lint to determine how many of each other type to use. This is described below the cross section table in more detail. |
|||||||||
Cross Section Tables: |
|||||||||
Dataset |
runN |
L=0 |
L=1 |
L=2 |
L=3 |
L=4 |
Total |
||
1 (Pt>10 GeV) e+e- |
1 |
262.6 |
94.0 |
32.3 |
11.18 |
n.a. |
400.0 |
||
1 (Pt>15 GeV) |
2+4 |
300.1 |
59.6 |
15.8 |
3.57 |
n.a. |
379.1 |
||
1 (Pt>20 GeV) |
3+7 |
325.5 |
40.4 |
9.0 |
1.45 |
n.a. |
376.4 |
||
|
|
|
|
|
|
|
|
||
2 (Pt>10 GeV) mu+mu- |
1 |
262.6 |
94.0 |
32.3 |
11.16 |
n.a. |
399.9 |
||
2 (Pt>15 GeV) |
2+4 |
300.1 |
59.6 |
15.8 |
3.56 |
n.a. |
379.1 |
||
2 (Pt>20 GeV) |
3+7 |
325.5 |
40.3 |
9.0 |
1.44 |
n.a. |
376.4 |
||
|
|
|
|
|
|
|
|
||
0 (Pt>10 GeV) e+ve |
1 |
667.2 |
320.2 |
113.2 |
29.7 |
10.5 |
1140.7 |
||
0 (Pt>15 GeV) |
2+4 |
801.0 |
218.0 |
55.5 |
10.6 |
2.25 |
1087.4 |
||
0 (Pt>20 GeV) |
3+7 |
876.4 |
153.4 |
31.0 |
4.6 |
0.67 |
1066.0 |
||
|
|
|
|
|
|
|
|
||
1 (Pt>10 GeV) mu+ vmu |
1 |
667.2 |
320.2 |
113.2 |
29.7 |
10.5 |
1140.7 |
||
1 (Pt>15 GeV) |
2+4 |
801.0 |
218.0 |
55.5 |
10.6 |
2.25 |
1087.4 |
||
1 (Pt>20 GeV) |
3+7 |
876.4 |
153.4 |
31.0 |
4.6 |
0.67 |
1066.0 |
||
|
|||||||||
Dealing with only Uniform Weight events: If you do not want to deal with weighted events, you can choose the event sample that corresponds to the smallest integrated luminosity, use ALL those events, and then use N = sigma * Lint to determine how many of each other type to use. Consider again the Z dataset "1" with Pt>10 GeV, i.e. 1.ZLrun1.M.io.
|
|||||||||
Another view on weighted events:
|
|||||||||
Physics Tidbits: Here is some important physics information that relates to the validity of the datasets:
|