My logs - Githubissues

ytatus94 commented 7 years ago

20170522
- Joey said use nBJet30_MV2c10 == 0 for b-jet veto
- Didn't see any possibility to calculate the filter efficiency.
- The FilterEffFromLog.py comes with SignalUncertaintiesUtils. It reads evgen.log to calculate filter efficiency and only supports Herwigpp sample.
- The SusySubprocessFinder must be included in the jobOptions. But our sample is MGPy8 and we didn't include SusySubprocessFinder.
- The module kilelib cannot be found so there is no way to run the FilterEffFromLog.py script.
- Sent a mail to the author of FilterEffFromLog.py.
- Sent a mail to Da Xu who is in charge of SUSYSignalUncertainties twiki page.

ytatus94 commented 7 years ago

20170523

Use AMI to list dataset ami list datasets --type EVNT mc15_13TeV.%.MGPy8EG_A14N23LO_NUHM2_m12_%_weak.%

which returns the matched patterns.

mc15_13TeV.370617.MGPy8EG_A14N23LO_NUHM2_m12_300_weak.evgen.EVNT.e4209
mc15_13TeV.370618.MGPy8EG_A14N23LO_NUHM2_m12_350_weak.evgen.EVNT.e4209
mc15_13TeV.370619.MGPy8EG_A14N23LO_NUHM2_m12_400_weak.evgen.EVNT.e4209
mc15_13TeV.370620.MGPy8EG_A14N23LO_NUHM2_m12_500_weak.evgen.EVNT.e4209
mc15_13TeV.370621.MGPy8EG_A14N23LO_NUHM2_m12_600_weak.evgen.EVNT.e4209
mc15_13TeV.370622.MGPy8EG_A14N23LO_NUHM2_m12_700_weak.evgen.EVNT.e4209
mc15_13TeV.370623.MGPy8EG_A14N23LO_NUHM2_m12_800_weak.evgen.EVNT.e4209

Use ami show dataset info: ami show dataset info mc15_13TeV.*.MGPy8EG_A14N23LO_NUHM2_m12_*_weak.evgen.EVNT.e4209
The results contain GenFiltEff_mean information.
The filter efficiency found using AMI

m12	crossSection_mean	GenFiltEff_mean
300	7.6337E-04	2.7338E-01
350	4.7445E-04	2.8948E-01
400	3.1222E-04	2.8592E-01
500	1.4145E-04	2.8348E-01
600	6.8266E-05	2.8765E-01
700	3.5047E-05	2.9353E-01
800	1.8791E-05	2.9840E-01

ytatus94 commented 7 years ago

20170517
- Filter efficiency
- xcalc_note.rtf and Baris_File_content.rtf
- prospino_sg_slha_200_300.txt and prospino_sg_slha_csca.dat
- outs_resdump and outs_sum.
- fn_resdump and outputbase_olddump.txt
- filter efficiency and effective cross-section strings
- --tev 13 --dir WeakSLHA --outdir WeakOut --batch --queue 2nd --process nn

ytatus94 commented 7 years ago

20170515
- Mike's slides have signal region definitions
- Shall we use the same SR definition?
  - Yes, we have to use the same SR definition as simplified Higgsino model.
- What is MTauTau?
  - Di-tau invariant mass squared calculated from MET and L1pT and L2pT
  - getMSqTauTau() is declared in SusySkimMaker/SusySkimMaker/Observables.h and implemented in SusySkimMaker/Root/Observables.cxx
  - There are 3 versions
  - version 1: arXiv:1401.1235, arXiv:1412.0618
  - version 2: arXiv:1409.7058, arXiv:1501.02511
  - version 3: CMS-PAS-SUS-16-025
- Mll use shape fit
- Where is the twiki page for Higgsino study?
- Joey said no twiki page but use JIRA tickets.
- Where can I download the common ntuple framework to produce the common ntuple?
- Joey said he can help us to produce the common ntuple.

ytatus94 commented 7 years ago

20170524

Using new METOverHT > max(5, 15-mll) cut, the number of events in the SR with and without cut are

sample	total entry	events (w/o METOverHT cut)	weighted events (w/o METOverHT cut)	events (w/ METOverHT cut)
300	5557	883.0	56.4883155823	0
350	7869	1333.0	35.5613250732	0
400	9958	1682.0	21.2211780548	0
500	13269	2709.0	6.99630832672	0
600	15051	3481.0	2.28578066826	1
700	15886	3814.0	0.645114243031	2
800	16600	4152.0	0.203541621566	3

All other cuts are applied.
Produce TRUTH3 derivation for Higgsino_160_150
The EVNT file is mc15_13TeV.393514.MGPy8EG_A14N23LO_SM_N2N1_160_150_2LMET50_MadSpin.evgen.EVNT.e5512
The outputs are put in raid05/Higgsino/TRUTH3_DAOD/mc15_13TeV.393514.MGPy8EG_A14N23LO_SM_N2N1_160_150_2LMET50_MadSpin.TRUTH3_DAOD.e5512

TTree Draw arguments with SR cuts:
- No METOverHT applied: "EventNumber","met_Et>200&&nJet30>0&&nJet30<5&&jetPt[0]>100&&nBJet30_MV2c10==0&&DPhiJ1Met>2.0&&(MTauTau<0||MTauTau>160)"
- With METOverHT applied: "met_Et>200&&nJet30>0&&nJet30<5&&jetPt[0]>100&&nBJet30_MV2c10==0&&DPhiJ1Met>2.0&&(MTauTau<0||MTauTau>160)&&(METOverHT>max(5, 15-mll))"

ytatus94 commented 7 years ago

20170526

Number of events in the SR with METOverHT cuts

Channel	events	weighted events
Wgamma	4677.0	5247.09570312
Wjets	227309.0	2616386.0
Zgamma	548.0	454.737854004
Zjets	34196.0	109052.890625
Zttgamma	801.0	1749.12353516
Zttjets	22170.0	15025.1962891
alt_DY_PowPy	12.0	2001.21533203
alt_ttbar_nonallhad	14700.0	2075375.375
diboson0L	984.0	179.619644165
diboson1L	21497.0	176822.234375
diboson2L	129733.0	17988.4414062
diboson3L	111388.0	1709.67932129
diboson4L	5931.0	5.93446874619
higgs	6761.0	42.4336853027
singletop	3541.0	10109.2783203
tZ	859.0	1.69478499889
topOther	159.0	0.00502032693475
triboson	1953.0	0.124958686531
ttV	8897.0	6.97038078308
ttbar	30566.0	197097.515625

ytatus94 commented 7 years ago

20170531

Joey said he forgot to use the absolute value for Dphi(lead jet, MET) so it kills all events with Dphi<-2.5. That is 45% of the events. And he used older b-tagging WP. So that could be the reason why my cutflow results are different. He will launch new production and check the cutflow results again.
NUHM2 m12 = 600 has negative weight

***** EventNumber=4304 RunNumber=999, RandomRunNumber=309640 eventWeight=1, pileupWeight=0.855144, genWeight=-1 leptonWeight=0.965636 bTagWeight=0.974268, jvtWeight=1.207 FS=86, xsec=-1
genWeight includes xsec.
If the SUSYTools cannot find the cross-section, then it returns -1.
NUHM2 doesn't have FS=86
Sent a mail to Joey to ask for help.
This only happens on m12=600. The other mass points are correct.
We should use METOverHTLep12 cut (not use METOverHT cut)
Number of events in SR with and without METOverHTLep12 cut

m12	events (w/o METOverHTLep12)	weighted events (w/o METOverHTLep12)	events (w/ METOverHTLep12)	weighted events (w/ METOverHTLep12)
300	172.0	64.6708755493	41.0	18.5641403198
350	281.0	63.6215248108	61.0	14.0987739563
400	349.0	57.1500968933	72.0	11.316160202
500	717.0	50.8394355774	157.0	10.475897789
600	905.0	32.1945648193	183.0	6.94837331772
700	1040.0	18.136680603	234.0	3.95941662788
800	1137.0	11.0251312256	219.0	2.31704163551

sample	events (w/o METOverHTLep12)	weighted events (w/o METOverHTLep12)	events (w/ METOverHTLep12)	weighted events (w/ METOverHTLep12)
Wgamma	42.0	6.57752847672	9.0	1.00712418556
Wjets	1537.0	117.365104675	719.0	61.613079071
Zgamma	51.0	0.672865927219	1.0	0.00427252845839
Zjets	1486.0	2.03024506569	102.0	13.2471857071
Zttgamma	50.0	0.91154718399	6.0	0.0558555871248
Zttjets	1529.0	47.4236564636	300.0	7.85560655594
alt_DY_PowPy	1.0	0.246106386185	0.0	0.0
alt_ttbar_nonallhad	1946.0	592.765136719	289.0	84.0288162231
diboson0L	1.0	0.0128046916798	1.0	0.0128046916798
diboson1L	642.0	361.702728271	121.0	47.3216552734
diboson2L	43234.0	491.895355225	5286.0	58.8162879944
diboson3L	23228.0	80.5093154907	2557.0	8.51855278015
diboson4L	1083.0	1.18642914295	100.0	0.0957641080022
higgs	1398.0	16.3626937866	378.0	3.53875637054
singletop	707.0	97.0341796875	78.0	10.3405408859
tZ	94.0	0.729139924049	12.0	0.094350501895
topOther	27.0	0.0737083926797	4.0	0.00971901696175
triboson	223.0	1.37124097347	29.0	0.162971973419
ttV	1276.0	5.0832195282	166.0	0.639597594738
ttbar	7494.0	538.792419434	894.0	60.6010856628

ytatus94 commented 7 years ago

20170601
- In HIGGSINO-19
- Joana: In particular the DeltaPhi cut is reverted for data and the DeltaR cut is reverted for the slepton signal sample.
- Jesse: genWeight * eventWeight * leptonWeight * jvtWeight * pileupWeight
  - genWeights = (xsec * filtEffic * lumi)/SumOfWeights
- Joey: The cosmic muon veto is no longer recommended, so we don't have it for the common ntuples. We don't yet have the bad muon veto for the common ntuples though, which I think should be added.

ytatus94 commented 7 years ago

20170606

Number of events in SR with METOverHT cuts:
The first two columns don't apply lep1Signal == 1 && lep2Signal ==1 requirement but the last two columns do apply this requirement.
From this table, with and without the lep1Signal == 1 && lep2Signal ==1 requirement have the same results.

sample	events	weighted events		events
m12_300	41.0	18.9524860382	41.0	18.9524860382
m12_350	61.0	14.0406732559	61.0	14.0406732559
m12_400	75.0	11.8051843643	75.0	11.8051843643
m12_500	160.0	10.6898288727	160.0	10.6898288727
m12_600	188.0	7.14116859436	188.0	7.14116859436
m12_700	238.0	4.04883813858	238.0	4.04883813858
m12_800	226.0	2.36664891243	226.0	2.36664891243
Higgsino_160_150	656.0	14.9409389496	656.0	14.9409389496
Wgamma	10.0	1.06457602978	10.0	1.06457602978
Wjets	733.0	64.9270172119	733.0	64.9270172119
Zgamma	1.0	0.00446981796995	1.0	0.00446981796995
Zjets	54.0	2.09397101402	54.0	2.09397101402
Zttgamma	6.0	0.0575345605612	6.0	0.0575345605612
Zttjets	303.0	8.08120822906	303.0	8.08120822906
alt_DY_PowPy	0.0	0.0	0.0	0.0
alt_ttbar_nonallhad	298.0	88.2742156982	298.0	88.2742156982
diboson0L	1.0	0.0130387032405	1.0	0.0130387032405
diboson1L	124.0	49.7006034851	124.0	49.7006034851
diboson2L	5413.0	61.3208808899	5413.0	61.3208808899
diboson3L	2603.0	8.80152893066	2603.0	8.80152893066
diboson4L	101.0	0.0976839289069	101.0	0.0976839289069
higgs	383.0	3.66811323166	383.0	3.66811323166
singletop	81.0	11.0275840759	81.0	11.0275840759
tZ	12.0	0.0946780368686	12.0	0.0946780368686
topOther	4.0	0.00984246004373	4.0	0.00984246004373
triboson	29.0	0.165587946773	29.0	0.165587946773
ttV	171.0	0.655358672142	171.0	0.655358672142
ttbar	917.0	64.7002410889	917.0	64.7002410889

Read output from SimpleAnalysis
- Use NUHM2 and Higgsino_160_150 to compare.
- Show the results in the meeting or put on JIRA

ytatus94 commented 7 years ago

20170607
- Meeting with Judita at 10am to discuss the MC production.
- Update the mll plot and made some comments about NUHM2 and send to Joey for EB meeting.
- Work on the internal note.

ytatus94 commented 7 years ago

20170619
- Add a Tau object and check the nTaus distributions
- Pat said use MinPt > 4
- Mike said the MadSpin doesn't affect the mll shape. We have to understand the reason why the mll is asymmetric for the NUHM2
- Pat said we use 2 lepton filter with MinPt > 2 for weak interaction
- Try to put the mass of C1 in the middle between N2 and N1 and check the mll shape.
- Z1SS = 1.37607727E+02
- -Z2SS = 1.59657379E+02
- W1SS = 1.53705612E+02
- Tomasso said Z2 and Z1 must be opposite sign.
- (Z1SS+Z2SS)/2 = 148.632553
- Unfortunately, the mll shape still the same.
- Study the reason why MTauTau has a peak at zero
- Set the initial value to -99999
- We reject events with met < 50 GeV before we declare the MTauTau. So the ntuple will fill zero for those events. That is why there is a peak at zero.
- Remove the met < 50 cut from the simple analysis framework and put this requirement when making plots.

ytatus94 commented 7 years ago

20170620
- There are a lot of events without electrons and muons. Why?
- Try to find the events with nBaselineLeptons == 0
- Compare nBaselineLeptons for with and without met < 50 cut cases
- ee, emu, mumu: nBaselineLeptons = 2
- etau, mutau: nBaselineLeptons = 1
- tautau: nBaselineLeptons = 1
- Some of the kinematic distributions have a peak at zero which might cause by the met < 50 rejection.
- Remove this met < 50 cut from the simple analysis and remake all the plots with the met cut.

ytatus94 commented 7 years ago

20170621
- Higgsino simplified model
- N2 --> N1 + Z, BR=0.9
- N2 --> C1+ + W-, BR=0.05
- N2 --> C1- + W+, BR=0.05
- C1+ --> N1 + W+, BR=1.0
- Mike suggests me to upload the MadGraph cards: run card, param card, and proc card. Also upload the control file to the JIRA.
- Remove the met < 50 rejection in the simple analysis framework.
- Add the met > 50 cut in the python script for the NUHM2 sample when make plots.

ytatus94 commented 7 years ago

20170622
- baseline
N El Mu Lep El+Mu

0 34229 33459 12679 67688

1 14359 14136 21357 28495

2 9995 10946 23619 20941

3 442 484 1820 926
- 442+484=926 < 1820
- The possible cases for nLep = 3 include nEl=3 + nMu=3 + nEl=2&&nMu=1 + nEl=1&&nMu=2
  - That is the reason why 926 < 1820 because 926 only contains nEl=3+nMu=3.
- MC
- Matrix element: MG5_aMC, Powheg, Alpgen,...
- Parton shower: Pythia8, Herwig++
- Combined: Sherpa, Herwig7
- Combinations: MG5_aMC+Py8, MG5_aMC+HWpp, Powheg+Py8, Powheg+HWpp
- After removing the met < 50 GeV rejection from the simple analysis framework, the mT2, HT30, and HTincl still have zero values.
- Add a requirement > 0 when make plots for these variables.

N	El	Mu	Lep	El+Mu
0	34229	33459	12679	67688
1	14359	14136	21357	28495
2	9995	10946	23619	20941
3	442	484	1820	926

ytatus94 commented 7 years ago

20170623

Prepare the python scripts to make masses and mass differences plots.
The same plots as Judita showed in her slides.
Compressed: N1, N2, C1
Accessible: N3, N4, C2, \tilde{g}, \tilde{t}, \bar{\tilde{t}}

N1	N2	N3	N4	C1	C2
N1	CC	CC	CA	CA	CC	CA
N2	CC	CC	CA	CA	CC	CA
N3	AC	AC	AA	AA	AC	AA
N4	AC	AC	AA	AA	AC	AA
C1	CC	CC	CA	CA	CC	CA
C2	AC	AC	AA	AA	AC	AA

Final state

FS
strong
2	gg (gluino - gluino)
61	tb1 (stop - antistop 1)
62	tb2 (stop - antistop 2)
weak
111	chi^{0}_1 - chi^{0}_1
112	chi^{0}_1 - chi^{0}_2
113	chi^{0}_1 - chi^{0}_3
114	chi^{0}_1 - chi^{0}_4
115	chi^{0}_1 - chi^{+}_1
116	chi^{0}_1 - chi^{+}_2
117	chi^{0}_1 - chi^{-}_1
118	chi^{0}_1 - chi^{-}_2
122	chi^{0}_2 - chi^{0}_2
123	chi^{0}_2 - chi^{0}_3
124	chi^{0}_2 - chi^{0}_4
125	chi^{0}_2 - chi^{+}_1
126	chi^{0}_2 - chi^{+}_2
127	chi^{0}_2 - chi^{-}_1
128	chi^{0}_2 - chi^{-}_2
133	chi^{0}_3 - chi^{0}_3
134	chi^{0}_3 - chi^{0}_4
135	chi^{0}_3 - chi^{+}_1
136	chi^{0}_3 - chi^{+}_2
137	chi^{0}_3 - chi^{-}_1
138	chi^{0}_3 - chi^{-}_2
144	chi^{0}_4 - chi^{0}_4
145	chi^{0}_4 - chi^{+}_1
146	chi^{0}_4 - chi^{+}_2
147	chi^{0}_4 - chi^{-}_1
148	chi^{0}_4 - chi^{-}_2
157	chi^{+}_1 - chi^{-}_1
158	chi^{+}_1 - chi^{-}_2
167	chi^{+}_2 - chi^{-}_1
168	chi^{+}_2 - chi^{-}_2

ytatus94 commented 7 years ago

20170626
- Download MG5_aMC_v2.5.5.tar and install it on my Mac.
- Read the MadGraph tutorial on indico
- Some useful reference of MadSpin
- website
- A slides found on indico

ytatus94 commented 7 years ago

20170627
- Chris' new result looks improve but he used generate p p > n2 x1+, n2 > l+ l- n1 add process p p > n2 x1+ j, n2 > l+ l- n1
- Read tutorial part 1 and part 2
- The new mll plot looks symmetric.
- Ask Chris to generate p p > n2 x1-, n2 > l+ l- n1
- Ask Chris to generate with and without jet

ytatus94 commented 7 years ago

20170628
- The new mll plots with and without jets look similar.
- Upload the mll plot w/o jet to JIRA
- Time needes:
- w/ jet: 1h40m
- w/o jet: 1h50m
- Try to produce MC by myself
- Generate_tf.py is located at
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.4/AtlasProduction/19.2.4.3/InstallArea/share/bin/Generate_tf.py
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.5/AtlasProduction/19.2.5.21/InstallArea/share/bin/Generate_tf.py
- compressed = n1 n2 x1- x1+
n1 n2 x1- x1+

n1 n1n1 n1n2 n1x1- n1x1+

n2 n2n1 n2n2 n2x1- n2x1+

x1- x1-n1 x1-n2 x1-x1- x1-x1+

x1+ x1+n1 x1+n2 x1+x1- x1+x1+

n1	n2	x1-	x1+
n1	n1n1	n1n2	n1x1-	n1x1+
n2	n2n1	n2n2	n2x1-	n2x1+
x1-	x1-n1	x1-n2	x1-x1-	x1-x1+
x1+	x1+n1	x1+n2	x1+x1-	x1+x1+

ytatus94 commented 7 years ago

20170629
- Download common ntuple v1.7 to ouhep04@cern and ouhep05@nhn
- ouhep04@cern: /UserDisk2/yushen/Higgsino/
- ouhep05@nhn: /raid05/atlas/data/Higgsino/
- Try to run Judita_N2C1p 10k
- Copy two root files from /afs/cern.ch/work/m/mamuzic/public/ForYuTing/NEW to
- Convert the root files to DAOD_TRUTH3

ytatus94 commented 7 years ago

20170703

Particle ID:

ID	particle		ID
1000022	N1
1000023	N2
1000024	C1
1	d (DN)	-1	DB
2	u (UP)	-2	UB
3	s (ST)	-3	SB
4	c (CH)	-4	CB
5	b (BT)	-5	BB
11	e- (E-)	-11	E+
12	Ve (NUE)	-12	ANUE
13	mu- (MU-)	-13	MU+
14	Vmu (NUM)	-14	ANUM
15	tau- (TAU-)	-15	TAU+
16	Vtau (NUT)	-16	ANUT

We need to produce N1N2, N2C1p, N2C1m, and C1C1
N1C1 has largest cross-section but only contains 1L.
- N1 doesn't decay
- C1+ -> W+ N1, W+ -> u \bar{d} or c \bar{s} or e \nu{e} or \mu \nu{\mu} or \tau \nu_{\tau}
N1N1 doesn't decay so we don't study this
N2N2 has smallest cross-section so we don't study this
N2 decay can be found from SLHA
Separate into 4 categories
N2 -> C1W: 420 to 429
N2 -> N1 q q: 430 to 434
N2 -> N1 l l: 435 to 437
N2 -> N1 v v: 438 to 440
Ignore N2 -> N1 gamma
We need to see the contribution from different N2 decay in N2C1p and N2C1m channels.
Use MadGraph only
run_07/: pp>n2 x1+-, n2>l+l-n1, x1+->ffn1
run_08/: pp>n2 x1+-, n2>vv~n1, x1+->ffn1
run_09/: pp>n2 x1+-, n2>qq~n1, x1+->ffn1
run_10/: pp>n2 x1+-, n2>ffx1+-, x1+->ffn1

ytatus94 commented 7 years ago

20170704
- Try to write a job option containing all the possible decays.
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.5/MCProd/19.2.5.14.3/InstallArea/python/MadGraphControl/MadGraphUtils.py
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.4/AtlasProduction/19.2.4.3/Generators/MadGraphControl/python/MadGraphUtils.py
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.5/AtlasProduction/19.2.5.14/InstallArea/share/bin/Generate_tf.py
- /cvmfs/atlas.cern.ch/repo/sw/software/x86_64-slc6-gcc47-opt/19.2.4/AtlasProduction/19.2.4.3/InstallArea/share/bin/Generate_tf.py

ytatus94 commented 7 years ago

20170705

Test the general MadGraphControl file
maxEvents and status

run	maxEvents	status		run	maxEvents	status
1	10k	12p	10k	16a	10k
2a	10k	fatal error	12m	10k	16b	10k
2b	10	13p	10k	16c	10	fatal error
2c	10	13m	10k	17a	10	fatal error
2d	10	14p	100	17b	10	fatal error
6	10k	14m	100	17c	10	fatal error
7	100	15p	100	18a	10	fatal error
8	100	15m	100	18b	10	fatal error
9	100	18c	10	fatal error
10	100	19a	10	fatal error
10	100	19b	10	fatal error
10	100	19c	10	fatal error

The fatal error messages:

generate 03:11:30 Py:Athena            INFO leaving with code 65: "failure in an algorithm execute"
generate 03:11:30 CORAL/Services/ConnectionService Info Deleting the ConnectionPool
PyJobTransforms.trfExe.execute 2017-07-05 03:11:30,146 INFO generate executor returns 65
PyJobTransforms.trfExe.validate 2017-07-05 03:11:30,202 ERROR Validation of return code failed: Non-zero return code from generate (65) (Error code 65)
PyJobTransforms.trfExe.validate 2017-07-05 03:11:30,213 INFO Scanning logfile log.generate for errors
PyJobTransforms.transform.execute 2017-07-05 03:11:30,360 CRITICAL Transform executor raised TransformValidationException: Non-zero return code from generate (65); Logfile err    or in log.generate: "Pythia8             FATAL ../src/GenModule.cxx:51 (StatusCode GenModule::execute()): code 0: this->callGenerator()"

This can be solved by increase the evt_multiplier value

ytatus94 commented 7 years ago

20170706

computers: using nproc to check number or cores and use less /proc/cpuinfo to check more information.

name	core	type	GHz
ouhep01	8	i7	3.60
ouhep02	4	i7	3.40
ouhep03	2	core 2	1.86
ouhep04	2	core 2	1.86
ouhep05	4	i7	3.40
ouhep06	2	core 2	3.16
ouhep07	4	i7	3.40
ouhep10	8	Xeon	1.86

Write DOE report
Fatal error: run_01, run_02a, run_16c, run_17a, run_17c, run_18c, run_19c
run_mg5.sh:

`run_16`	`run_17`	`run_18`	`run_19`
MadSpin	10k, 19.2.5.14.3	10, 19.2.5.14.13	10k, 19.2.5.14.3	10k, 19.2.5.14.3
MadGraph	10k, 19.2.5.14.3	10, 19.2.5.14.13	10, 19.2.5.14.3	10k, 19.2.5.14.3
Pythia	10, 19.2.5.14.3	10, 19.2.5.14.13	10k, 19.2.5.14.3	10k, 19.2.5.14.3

Shift training exam
Apply scholarship
Flight ticket to Portugal
Ture ACA form
Write DOE report (software, talk, paper, shift)

ytatus94 commented 7 years ago

20170707

Discussed with Judita at 10:40am in R1
Run N2N1, C1C1, N2C1p, N2C1m using Judita's jobOption.
The use_syst1, ktdurham, maxjetflaver, sys_alpsfact, ickkw values in my job option and in Judita's job option are different.
ktdurham is removed
Pat said we should use MLM, so ickkw = 1
Convert 12p, 12m, 13p, 13m, 14p, 14m, 15p, 15m to TRUTH3

run	decay	weighted filter efficiency	calculation
12p	p p > n2 x1+, n2 > l+ l- n1, x1+ > f f n1	0.605510	10000.000000 / 16515.000000
12m	p p > n2 x1-, n2 > l+ l- n1, x1- > f f n1	0.619579	10000.000000 / 16140.000000
13p	p p > n2 x1+, n2 > v v~ n1, x1+ > f f n1	0.009555	1051.000000 / 109993.000000
13m	p p > n2 x1-, n2 > v v~ n1, x1- > f f n1	0.010037	1104.000000 / 109993.000000
14p	p p > n2 x1+, n2 > q q~ n1, x1+ > f f n1	0.029174	3209.000000 / 109997.000000
14m	p p > n2 x1-, n2 > q q~ n1, x1- > f f n1	0.028922	3181.000000 / 109985.000000
15p	p p > n2 x1+, n2 > f f x1, x1+ > f f n1	0.117129	10000.000000 / 85376.000000
15m	p p > n2 x1-, n2 > f f x1, x1- > f f n1	0.123213	10000.000000 / 81160.000000

We assume that the BR(pp > n2 X1+) and BR(pp > n2 X1-) are 100% in MadGraph.
The BR(X1+- > ff N1) is 100%
But in MadGraph, the BR(n2 > l+ l- n1), BR(n2 > v v~ n1), BR(n2 > q q~ n1), BR(n2 > f f n1) are also assumed to be 100%.
In the reality, the BRs are not 100%, they are

Decay BR

n2 > f f n1 0.00447339029

n2 > l+ l- n1 0.1069829426

n2 > q q~ n1 0.6654233643

n2 > v v~ n1 0.2192035690
MadGraphControl_SimplifiedModel_higgsino.py
MadGraphControl_SimplifiedModelPreInclude.py

Decay	BR
n2 > f f n1	0.00447339029
n2 > l+ l- n1	0.1069829426
n2 > q q~ n1	0.6654233643
n2 > v v~ n1	0.2192035690

ytatus94 commented 7 years ago

20170710
- Prepare slides to present in the Higgsino LSP meeting
- That is what I have for introduction
- very promising
- for quick reference
- moving on slide 4
- ok, that's everything
- It just to me...
- for my perspective
- one of the concern I have is
- Try to run Judita's new JO
- \hyperset{linkbordercolor=black}

ytatus94 commented 7 years ago

20170711
- Run 10k events and check the filter efficiency
production filter efficiency weighted filter efficiency

N2N1 0.304016 0.331336

C1C1 0.285250 0.303839

N2C1p 0.301996 0.325035

N2C1m 0.301150 0.323453
- Run 50k events and check the filter efficiency and then run the analysis to see how many events exist in the SR.
- This can help us to determine how many events we have to request.
- Higgsino simplified model N2=170 and N1=150:
- list of dataset
```
mc15_13TeV.393515.MGPy8EG_A14N23LO_SM_N2N1_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393564.MGPy8EG_A14N23LO_SM_C1C1_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393417.MGPy8EG_A14N23LO_SM_N2C1p_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393466.MGPy8EG_A14N23LO_SM_N2C1m_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
```
- EVNT files:
EVNT files

N2N1 09999255_00001

09999255_00002

C1C1 09999559_00004

09999255_00005

N2C1p 09998720_00001

09998720_00002

N2C1m 09998992_00001

09998992_00002

production	filter efficiency	weighted filter efficiency
N2N1	0.304016	0.331336
C1C1	0.285250	0.303839
N2C1p	0.301996	0.325035
N2C1m	0.301150	0.323453

EVNT files
N2N1	09999255_00001
09999255_00002
C1C1	09999559_00004
09999255_00005
N2C1p	09998720_00001
09998720_00002
N2C1m	09998992_00001
09998992_00002

ytatus94 commented 7 years ago

20170712

Higgsino simplified model

DSID	production	FS	Xsec	BF	Filter Eff
393417	N2C1p_170_150	125	0.578138	0.109839	1.704238e-01
393466	N2C1m_170_150	127	0.328355	0.109839	1.668405e-01
393515	N2N1_170_150	112	0.562441	0.109839	1.582031e-01
393564	C1C1_170_150	157	0.563417	0.110958	7.314584e-02

n2 > l+ l- n1: BR = 9.00000000E-01
c1+ > f f n1: BR = 9.42923324E-02

ytatus94 commented 7 years ago

20710714
- run_24, run_25, run_26, run_27 is running on ouhep01
- Use 2L multiplier 100, MET multiplier 10
- Use MET50
- 10k events

ytatus94 commented 7 years ago

20170716
- Entries
decay Higgsino entries nuhm2 entries

N2N1 10000 10000

C1C1 10000 10000

N2C1p 10000 10000

N2C1m 10000 10000
- Filter efficiency
decay efficiency weight efficiency

N2N1 0.165175 0.178518

C1C1 0.107170 0.114324

N2C1p 0.175818 0.189159

N2C1m 0.175864 0.188560

decay	Higgsino entries	nuhm2 entries
N2N1	10000	10000
C1C1	10000	10000
N2C1p	10000	10000
N2C1m	10000	10000

decay	efficiency	weight efficiency
N2N1	0.165175	0.178518
C1C1	0.107170	0.114324
N2C1p	0.175818	0.189159
N2C1m	0.175864	0.188560

ytatus94 commented 7 years ago

20170717
- Number of events in SR
- Both Higgsino and NUHM2 generate 10k events
- Use old SR definitions
decay Higgsino NUHM2

N2N1 557 500

C1C1 380 325

N2C1p 474 414

N2C1m 477 401
- Use new SR definitions
decay Higgsino NUHM2

N2N1 646 586

C1C1 390 331

N2C1p 537 488

N2C1m 536 468
- Take after shift and write thesis

decay	Higgsino	NUHM2
N2N1	557	500
C1C1	380	325
N2C1p	474	414
N2C1m	477	401

decay	Higgsino	NUHM2
N2N1	646	586
C1C1	390	331
N2C1p	537	488
N2C1m	536	468

ytatus94 commented 7 years ago

20170721
- run 7 m12 points and 4 decays on
m12 ouhep m12 ouhep

300 06 600 01

350 03 700 01

400 07 800 01

500 07
- use Judita's new JO uploaded to JIRA
- produce 20k for each sample.

m12	ouhep		m12	ouhep
300	06	600	01
350	03	700	01
400	07	800	01
500	07

ytatus94 commented 7 years ago

20170722

Run on local machine

m12	N2N1	C1C1	N2C1p	N2C1m
300	?	FATAL	ok	ok
350	?	?	exit code 137	exit code 137
400	ok	FATAL	ok	ok
500	ok	FATAL	ok	ok
600	ok	FATAL	ok	FATAL
700	FATAL	FATAL	ok	FATAL
800	FATAL	FATAL	ok	ok

Time

m12	N2N1	C1C1	N2C1p	N2C1m
300	1454m22.727s	1048m54.233s	801m38.493s	759m30.919s
350	1448m51.128s	1451m48.203s	996m8.022s	996m22.700s
400	ok	FATAL	ok	ok
500	ok	FATAL	ok	ok
600	917m17.878s	701m52.840s	564m56.084s	464m42.912s
700	901m10.938s	733m43.805s	636m12.517s	522m42.864s
800	888m59.880s	697m27.459s	636m40.764s	588m28.072s

Filter efficiency (multiplier = 100 for 2L3 and 20 for MET50)

m12	N2N1	C1C1	N2C1p	N2C1m
300	0.150295 [19020 / 126551]	0.259946 [20000 / 76939]	0.262892 [20000 / 76077]
350
400	0.212823 [20000 / 93975]	0.127010 [16372 / 128903]	0.228728 [20000 / 87440]	0.229822 [20000 / 87024]
500	0.181220 [20000 / 110363]	0.116842 [16692 / 142860]	0.193515 [20000 / 103351]	0.195461 [20000 / 102322]
600	0.164974 [20000 / 121231]	0.107088 [15784 / 147393]	0.177252 [20000 / 112834]	0.175867 [18998 / 108025]
700	0.147835 [17102 / 115683]	0.100172 [17221 / 171914]	0.160754 [20000 / 124414]	0.163473 [16184 / 99001]
800	0.134384 [14501 / 107907]	0.093320 [12962 / 138898]	0.150851 [20000 / 132581]	0.151368 [20000 / 132128]

ytatus94 commented 6 years ago

20170724

Filter efficiency (multiplier = 100 for 2L3 and 50 for MET50)

m12	N2N1	C1C1	N2C1p	N2C1m
300	0.404065 [20000 / 49497]	0.149810 [19457 / 129878]	0.261417 [20000 / 76506]	0.261134 [20000 / 76589]
350	0.236754 [20000 / 84476]	0.137365 [20000 / 145597]	0.261931 [20000 / 76356]	0.258662 [20000 / 77321]
400	0.213465 [20000 / 93692]	0.127532 [15963 / 125169]	0.230282 [20000 / 86850]	0.230165 [20000 / 86894]
500	0.184140 [20000 / 108613]	0.117058 [16698 / 142647]	0.194534 [20000 / 102810]	0.193966 [20000 / 103111]
600	0.163818 [20000 / 122087]	0.106505 [15694 / 147355]	0.174724 [20000 / 114466]	0.175718 [19029 / 108293]
700	0.147799 [16997 / 115001]	0.099473 [17098 / 171885]	0.161406 [20000 / 123911]	0.164326 [16321 / 99321]
800

ytatus94 commented 6 years ago

20170725

Muhammad says that using evgenConfig.minevents = can solve the fatal error code 0: this->callGenerator().
But it only works on local machine. MC production team doesn't allow we do this.

Filter efficiency:

m12	N2N1	C1C1	N2C1p	N2C1m
300	0.404065 [20000 / 49497]	0.150175 [19702 / 131194]	0.261417 [20000 / 76506]	0.261134 [20000 / 76589]
350	0.236754 [20000 / 84476]	0.136515 [20000 / 146504]	0.261931 [20000 / 76356]	0.258662 [20000 / 77321]
400	0.213465 [20000 / 93692]]	0.129076 [16799 / 130148]	0.230282 [20000 / 86850]	0.230165 [20000 / 86894]
500	0.184140 [20000 / 108613]	0.115900 [16564 / 142916]	0.194534 [20000 / 102810]	0.193966 [20000 / 103111]
600	0.161633 [20000 / 123737]	0.106692 [15600 / 146215]	0.174940 [20000 / 114325]	0.177813 [19028 / 107011]
700	0.147311 [16949 / 115056]	0.099933 [17197 / 172085]	0.163107 [20000 / 122619]	0.162471 [16082 / 98984]
800	0.132845 [14339 / 107938]	0.095604 [13900 / 145392]	0.149621 [20000 / 133671]	0.150488 [20000 / 132901]

The denominator in the filter efficiency calculation.
The values are similar even we increase the multiplier.

m12	multiplier	N2N1	C1C1	N2C1p	N2C1m
300	20	-	-	76939	76077
300	50	49497	126551	76506	76589
300	80	-	131194	76506	76589
350	20	-	-	-	-
350	50	84476	145597	76356	77321
350	80	-	146504	-	-
400	20	93975	128903	87440	87024
400	50	93692	125169	86850	86894
400	80	-	130148	-	-
500	20	110363	142860	103351	102322
500	50	108613	142647	102810	103111
500	80	-	142916	-	-
600	20	121231	147393	112834	108025
600	50	122087	147355	114466	108293
600	80	123737	146215	114325	107011
700	20	115683	171914	124414	99001
700	50	115001	171885	123911	99321
700	80	115056	172085	122619	98984
800	20	107907	138898	132581	132128
800	50	107938	145392	133671	132901
800	80	-	-	-	-

extension of the dataset

ytatus94 commented 6 years ago

20170726
- Ximo said: About the acceptance, it corresponds to the fraction of events passing your selection cuts at truth level
- acceptance: N(pass truth)/N(all)
- efficiency: N(pass reco)/N(pass truth)
- acceptance x eff: N(pass reco)/N(all)

ytatus94 commented 6 years ago

20170727
- Peter Tornambe:
- git repo: https://gitlab.cern.ch/disimone/SSHistHitter
- Ntuple code: svn co svn+ssh://$USER@svn.cern.ch/reps/atlas-fcardill/DiLepton DiLepton
- /afs/cern.ch/user/p/ptornamb/public/DiLepton

ytatus94 commented 6 years ago

20170728

multiplier=100 (2L3), 20(MET50), --maxEvents=5000, JO 20170719 + 19.2.5.21.1

m12		N2N1	C1C1	N2C1p	N2C1m
300	eff	0.409333 [5000 / 12215]	0.152448 [5000 / 32798]	0.253807 [5000 / 19700]	0.265449 [5000 / 18836]
time	1449	982	755	716
350	eff	0.233383 [5000 / 21424]	0.137159 [5000 / 36454]	0.260173 [5000 / 19218]	0.261520 [5000 / 19119]
time	1291	1007	762	687
400	eff	0.214160 [5000 / 23347]	0.129604 [5000 / 38579]	0.228425 [5000 / 21889]	0.233951 [5000 / 21372]
time	1282	993	806	717
500	eff	0.183824 [5000 / 27200]	0.115923 [5000 / 43132]	0.191344 [5000 / 26131]	0.194447 [5000 / 25714]
time	1004	800	724
600	eff	0.163084 [5000 / 30659]	0.106943 [5000 / 46754]	0.170091 [5000 / 29396]	0.174929 [5000 / 28583]
time	822	400	284	193
700	eff	0.147171 [5000 / 33974]	0.097254 [5000 / 51412]	0.161943 [5000 / 30875]	0.164026 [5000 / 30483]
time	803	487	321	273
800	eff	0.135208 [5000 / 36980]	0.094652 [5000 / 52825]	0.149589 [5000 / 33425]	0.154617 [5000 / 32338]
time	683	433	342	301

multiplier=100 (2L3), 20(MET50), --maxEvents=5000, JO 20170726 + 19.2.5.21.1

m12		N2N1	C1C1	N2C1p	N2C1m
600	eff	0.162377 [2000 / 12317]	0.101874 [500 / 4908]	0.176960 [2000 / 11302]	0.170809 [2000 / 11709]
time	696	461	402	325
700	eff	0.146875 [2000 / 13617]	0.096993 [500 / 5155]	0.169233 [2000 / 11818]	0.161434 [2000 / 12389]
time	691	466	395	305
800	eff	0.136454 [2000 / 14657]	0.094949 [500 / 5266]	0.149109 [2000 / 13413]	0.151757 [2000 / 13179]
time	691	463	391	332

ytatus94 commented 6 years ago

20170730
- Number of events in SR
- TRUTH3
  
  m12 N2N1 C1C1 N2C1p N2C1m
  
  300 117 123 146 129
  
  350 197 137 129 139
  
  400 221 135 197 172
  
  500 266 171 234 200
  
  600 281 160 249 246
  
  700 303 185 234 229
  
  800 298 180 275 233

m12	N2N1	C1C1	N2C1p	N2C1m
300	117	123	146	129
350	197	137	129	139
400	221	135	197	172
500	266	171	234	200
600	281	160	249	246
700	303	185	234	229
800	298	180	275	233

ytatus94 commented 6 years ago

20170731

Weighted filter efficiency

m12		N2N1	C1C1	N2C1p	N2C1m
300	eff	0.414405 [6496.438761 / 15676.546351]	0.164017 [7001.824881 / 42689.710359]	0.265627 [7045.985973 / 26525.874342]	0.277531 [7069.025200 / 25471.104999]
350	eff	0.248245 [6981.815608 / 28124.703346]	0.148728 [7081.169635 / 47611.689525]	0.273310 [7100.283196 / 25978.823885]	0.275627 [7127.214925 / 25858.159898]
400	eff	0.231079 [7110.087579 / 30769.121271]	0.141000 [7118.357039 / 50484.913969]	0.244031 [7194.425192 / 29481.576529]	0.249081 [7210.782080 / 28949.496191]
500	eff	0.198565 [7091.717201 / 35714.895181]	0.125613 [7096.000634 / 56490.845087]	0.206531 [7291.280653 / 35303.548179]	0.207577 [7240.862591 / 34882.698452]
600	eff	0.177126 [7149.202958 / 40362.284916]	0.115093 [7081.053773 / 61524.655917]	0.182043 [7201.247888 / 39557.933896]	0.188051 [7306.759687 / 38855.206895]
700	eff	0.160043 [7140.607424 / 44616.914332]	0.103832 [7011.838356 / 67530.508958]	0.174270 [7254.922130 / 41630.438908]	0.176030 [7300.162934 / 41471.058299]
800	eff	0.146843 [7142.989645 / 48643.557850]	0.101354 [7028.658871 / 69347.365880]	0.161460 [7249.957357 / 44902.393304]	0.165694 [7274.957122 / 43905.932712]

Acceptance = N_SR / weighted pass truth events

m12	N2N1	C1C1	N2C1p	N2C1m
300	0.01800986729873986	0.017566848941590947	0.02072101769141572	0.018248626415987314
350	0.028216156235101764	0.019347086295299576	0.018168289410297488	0.019502709187628434
400	0.0310825988490964	0.018965050398619097	0.02738231265773095	0.023853168504018914
500	0.03750854587976117	0.024098081274213146	0.032093127550058116	0.02762101855773222
600	0.03930508081121957	0.02259550698655597	0.03457734046552238	0.03366745459518491
700	0.04243336483974728	0.026383951056386846	0.032253964385417874	0.03136916286257783
800	0.04171922609584015	0.02560943749065326	0.037931257586567894	0.032027680176339605

N_SR normalize to lumi=36.1 fb-1

m12	N2N1	C1C1	N2C1p	N2C1m	Total
300	0.028181674341720638	13.425648881262498	6.752121505656585	3.628274430453513	23.8342264917
350	15.787892204305011	10.99886932538464	8.729224265740894	5.508887433191271	41.0248732286
400	17.35553599231772	9.008782796224596	11.767022150713228	6.078409972752035	44.209750912
500	18.996405926778642	8.822490107795607	11.627697282292507	5.83484703369516	45.2814403506
600	17.90062660660316	7.021948921561087	10.997488704486194	6.401584699963968	42.3216489326
700	17.426776470585377	7.061130715714346	9.783797615295533	5.5589907838062995	39.8306955854
800	15.613511005888435	6.484006342223556	10.632776449955497	5.332588033465056	38.0628818315

ytatus94 commented 6 years ago

20170801

BR

m12	BR	N2N1	C1C1	N2C1p	N2C1m
300	BR(N2->llN1)	0.0378120923	x	0.0378120923	0.0378120923
BR(C1->ffN1)	x	x	0.666666806	0.666666806
BR(C1->lvlN1)	x	0.333333194	x	x
350	BR(N2->llN1)	0.1013306826	x	0.1013306826	0.1013306826
BR(C1->ffN1)	x	x	0.66666472	0.66666472
BR(C1->lvlN1)	x	0.33333601	x	x
400	BR(N2->llN1)	0.1028803438	x	0.1028803438	0.1028803438
BR(C1->ffN1)	x	x	0.666664064	0.666664064
BR(C1->lvlN1)	x	0.33333601	x	x
500	BR(N2->llN1)	0.1051690615	x	0.1051690615	0.1051690615
BR(C1->ffN1)	x	x	0.666664064	0.666664064
BR(C1->lvlN1)	x	0.333335965	x	x
600	BR(N2->llN1)	0.1069829426	x	0.1069829426	0.1069829426
BR(C1->ffN1)	x	x	0.666664362	0.666664362
BR(C1->lvlN1)	x	0.333335683	x	x
700	BR(N2->llN1)	0.1084250846	x	0.1084250846	0.1084250846
BR(C1->ffN1)	x	x	0.66666454	0.66666454
BR(C1->lvlN1)	x	0.333335459	x	x
800	BR(N2->llN1)	0.1094692499	x	0.1094692499	0.1094692499
BR(C1->ffN1)	x	x	0.66666466	0.66666466
BR(C1->lvlN1)	x	0.333335274	x	x

Acceptance = N_SR / weighted pass truth events

m12	N2N1	C1C1	N2C1p	N2C1m
300	0.01800986729873986	0.017566848941590947	0.02072101769141572	0.018248626415987314
350	0.028216156235101764	0.019347086295299576	0.018168289410297488	0.019502709187628434
400	0.0310825988490964	0.018965050398619097	0.02738231265773095	0.023853168504018914
500	0.03750854587976117	0.024098081274213146	0.032093127550058116	0.02762101855773222
600	0.03930508081121957	0.02259550698655597	0.03457734046552238	0.03366745459518491
700	0.04243336483974728	0.026383951056386846	0.032253964385417874	0.03136916286257783
800	0.04171922609584015	0.02560943749065326	0.037931257586567894	0.032027680176339605

N_SR normalize to lumi=36.1 fb-1

m12	N2N1	C1C1	N2C1p	N2C1m	Total
300	0.009958953938109184	13.452529263419953	2.388480957152261	1.2834580031180856	17.1344271776
350	14.951382185770518	11.021077101060241	8.274963473420229	5.2222100041294315	39.4696327644
400	16.687322520774963	9.026972395676154	11.325255706207159	5.850209708698811	42.8897603314
500	18.67134750647054	8.840301177518182	11.440124577705426	5.740721944918655	44.6924952066
600	17.897772979049147	7.036113107447679	11.006704232312819	6.406949013901654	42.3475393327
700	17.658875823630925	7.07536442715252	9.92399596013507	5.638649147308162	40.2968853582
800	15.973825589906555	6.497069484111924	10.889006378389656	5.461093382619587	38.820994835

ytatus94 commented 6 years ago

20170802

Mike said using 160_100, 140_100, and 170_150
mc15_13TeV.393510.MGPy8EG_A14N23LO_SM_N2N1_160_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393559.MGPy8EG_A14N23LO_SM_C1C1_160_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393412.MGPy8EG_A14N23LO_SM_N2C1p_160_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393461.MGPy8EG_A14N23LO_SM_N2C1m_160_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393509.MGPy8EG_A14N23LO_SM_N2N1_140_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393558.MGPy8EG_A14N23LO_SM_C1C1_140_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393411.MGPy8EG_A14N23LO_SM_N2C1p_140_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393460.MGPy8EG_A14N23LO_SM_N2C1m_140_100_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393515.MGPy8EG_A14N23LO_SM_N2N1_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393564.MGPy8EG_A14N23LO_SM_C1C1_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393417.MGPy8EG_A14N23LO_SM_N2C1p_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
mc15_13TeV.393466.MGPy8EG_A14N23LO_SM_N2C1m_170_150_2LMET50_MadSpin.evgen.EVNT.e5512
Higgsino sample status

n2_n1	Download EVNT	Convert to TRUTH3	Run SimpleAnalysis
160_100	V	V	V
140_100	V	V	V
170_150	V	V	V

NUHM2 N2 and N1 masses

m12	N2	N1	dM
300	161.3078	105.631409	55.676391
350	161.675125	115.618103	46.057022
400	161.1427	122.974541	38.168159
500	160.295654	132.279099	28.016555
600	159.657379	137.607727	22.049652
700	159.170868	140.978836	18.192032
800	158.783905	143.288208	15.495697

ytatus94 commented 6 years ago

20170817

DSID

Higgsino	N2N1	C1C1	N2C1p	N2C1m
140_100	393509	393558	393411	393460
160_100	393510	393559	393412	393461
170_150	393515	393564	393417	393466
190_150	393516	393565	393418	393467

In truth sample, the weight = cross-section x filter efficiency x luminosity x event weight / total event weight
No lepton weight, jvt weight, and pileup weight in the truth.
In simple analysis
acceptance = sum of event weights passing SR / sum of all event weights
N_{event} = acceptance x cross-section x luminosity
There is a branch call eventWeight and this is the mc event weight.
The histograms produced by simple analysis already consider event weight. We just need to normalize it to the total number of weighted events and then scale it up again. i.e. [hist(already consider event weight) / sum of all event weight] x cross-section x filter efficiency x luminosity

Judita's code snippet

hev = TH1F("ev","ev",1000,-1,100000000)
ntuple.Draw("Event>>+ev","eventWeight") 
events = hev.Integral()
entries = ntuple.GetEntries()
weight = xsecs[dsid]["xsec"] * xsecs[dsid]["kfac"] * xsecs[dsid]["gfe"] * LUMI * 1000 / events
accept
f = openFile(mergedir+"/"+str(ds)+".root")
ntuple = f.Get("ntuple")
afname = mergedir+"/"+str(ds)+".txt"
accres = readAcceptanceFile(afname)
hev = TH1F("ev","ev",1000,-1,100000000)
ntuple.Draw("Event>>+ev","eventWeight") 
events = hev.Integral()
entries = hev.GetEntries()
print "... Events", events, "histogram entries", hev.GetEntries(), "simple analysis entries", float(accres["All"]["events"])
if int(hev.GetEntries()) != int(float(accres["All"]["events"])): print " WARNING: Possible problem in merging for sample", dsid
weight = xsecs[ds]["xsec"] * xsecs[ds]["kfac"] * xsecs[ds]["gfe"] * LUMI * 1000 / entries
xsecs[dsid] = {"physics": phys, "xsec": xsec, "kfac": kfac, "gfe": gfe, "relunc": runc}
def readAcceptanceFile(fname):
#print "Read acceptance for", fname
f = open(fname)
contents = f.readlines()
keys = (contents[0][:-1]).split(",")
accres = {}
count = 0
for line in contents:
info = line[:-1].split(",")
if count != 0:
#print info[0], keys[0]
accres[info[0]] = {keys[1]: info[1], keys[2]: info[2],keys[3]: info[3]}
count = count + 1
#for acc in accres:
# print acc, accres[acc]
return accres

ytatus94 commented 6 years ago

20170822

Number of events in SR

sample	N2N1	C1C1	N2C1p	N2C1m
140_160	420	279	338	311
160_100	306	251	217	198
170_150	646	390	537	536
190_150	518	377	423	371
m12=300	117	123	146	129
m12=350	197	137	129	139
m12=400	221	135	197	172
m12=500	266	171	234	200
m12=600	281	160	249	246
m12=700	303	185	234	229
m12=800	298	180	275	233

ytatus94 commented 6 years ago

20170823

Adding a branch to an existing tree

void upd() {
TFile *f = new TFile("hs.root","update");
TTree *T = (TTree*)f->Get("ntuple");
float px,py;
float pt;
TBranch *bpt = T->Branch("pt",&pt,"pt/F");
T->SetBranchAddress("px",&px);
T->SetBranchAddress("py",&py);
Long64_t nentries = T->GetEntries();
for (Long64_t i=0;i<nentries;i++) {
  T->GetEntry(i);
  pt = TMath::Sqrt(px*px+py*py);
  bpt->Fill();
}
T->Print();
T->Write();
delete f;
}

To add new output to an existing Tree, do:
open old file in "update" mode
Get the Tree from the file
```
TTree *tree = (TTree*)f.Get("tree");
```
set tree branch address(es)
fill
tree->Write()
close file

copytree.C

1 /// \file
2 /// \ingroup tutorial_tree
3 /// \notebook -nodraw
4 /// Copy a subset of a Tree to a new Tree
5 ///
6 /// The input file has been generated by the program in `$ROOTSYS/test/Event`
7 /// with `Event 1000 1 1 1`.
8 ///
9 /// \macro_code
10 ///
11 /// \author Rene Brun
12 
13 R__LOAD_LIBRARY($ROOTSYS/test/libEvent.so)
14 
15 void copytree() {
16    //Get old file, old tree and set top branch address
17    TFile *oldfile;
18    TString dir = "$ROOTSYS/test/Event.root";
19    gSystem->ExpandPathName(dir);
20    if (!gSystem->AccessPathName(dir))
21        {oldfile = new TFile("$ROOTSYS/test/Event.root");}
22    else {oldfile = new TFile("./Event.root");}
23    TTree *oldtree = (TTree*)oldfile->Get("T");
24    Event *event   = new Event();
25    oldtree->SetBranchAddress("event",&event);
26    oldtree->SetBranchStatus("*",0);
27    oldtree->SetBranchStatus("event",1);
28    oldtree->SetBranchStatus("fNtrack",1);
29    oldtree->SetBranchStatus("fNseg",1);
30    oldtree->SetBranchStatus("fH",1);
31 
32    //Create a new file + a clone of old tree in new file
33    TFile *newfile = new TFile("small.root","recreate");
34    TTree *newtree = oldtree->CloneTree();
35 
36    newtree->Print();
37    newfile->Write();
38    delete oldfile;
39    delete newfile;
40 }

Root tutorial copytree.C, copytree2.C, and copytree3.C

ytatus94 commented 6 years ago

20170825
- The cross section of pp->N2N1 is \sigma and the LHC integrate luminosity is \int \mathcal{L}dt
- Nevents(pp->N2N1) = \sigma \times \int \mathcal{L}dt
- The branching ratio for N2 ->\ell\ell N1 is BR
  - Nevents(pp->N2N1->(\ell\ell N1)N1) = (\sigma \times \int \mathcal{L}dt) \times BR
- There are 2L3 and MET50 filters and the filter efficiency is \epsilon_{filter}
  - Nevents(pass filter) = Nevents(pp->N2N1->(\ell\ell N1)N1) \times \epsilon_{filter} pass the filter
- In the signal region, the acceptance is A, the total number of events generated and pass filters and pass the SR selection requirements are
  - Nevents(final) = Nevents(pass filter) \times A = Nevents(pp->N2N1->(\ell\ell N1)N1) \times \epsilon{filter} \times A = (\sigma \times \int \mathcal{L}dt) \times BR \times \epsilon{filter} \times A

ytatus94 commented 6 years ago

20170831
- Use hadd to produce new signal sample.
- Run exclusion fit for NUHM2:
- HistFitter.py -t -w -f -D "corrMatrix" -F excl -g 'MGPy8EG_A14N23LO_SM_Higgsino_100_80_2LMET50_MadSpin' $HISTFITTER/analysis/higgsino/HiggsinoConfig.py
- HistFitter.py -F excl -l -g 'MGPy8EG_A14N23LO_SM_Higgsino_100_80_2LMET50_MadSpin' $HISTFITTER/analysis/higgsino/HiggsinoConfig.py
- Run --NoSys first
- In HiggsinoFitter/contours/hypotest_to_csv.sh, follow all steps related Higgsino_all_points_upperlimit.root, than makelistfiles.C()

ytatus94 commented 6 years ago

20170905

Use Mike's SusyHit file, the new BR's are:

m12	BR(N2->llN1)	BR(C1+/- ->ffN1)	BR(C1->lvlN1)
susyhit_slha.370618.out	350	0.1013847148	0.666742752	0.333257248
susyhit_slha.370619.out	400	0.1029349381	0.666762124	0.333237877
susyhit_slha.370620.out	500	0.105384522	0.666804296	0.333195704
susyhit_slha.370621.out	600	0.1076035524	0.666841171	0.333158828
susyhit_slha.370622.out	700	0.1097007757	0.666868795	0.333131205
susyhit_slha.370623.out	800	0.1116429166	0.666887568	0.333112432

ytatus94 commented 6 years ago

20170923

Check the filter efficiency form AMI for the new NUHM2 MC samples --> Done!
Use AMI to list dataset ami list datasets --type EVNT mc15_13TeV.%.MGPy8EG_A14N23LO_NUHM2_m12_%_weak_%2LMET50_MadSpin%
Return the matched pattern mc15_13TeV.394301.MGPy8EG_A14N23LO_NUHM2_m12_300_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394302.MGPy8EG_A14N23LO_NUHM2_m12_300_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394303.MGPy8EG_A14N23LO_NUHM2_m12_300_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394304.MGPy8EG_A14N23LO_NUHM2_m12_300_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394305.MGPy8EG_A14N23LO_NUHM2_m12_350_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394306.MGPy8EG_A14N23LO_NUHM2_m12_350_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394307.MGPy8EG_A14N23LO_NUHM2_m12_350_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394308.MGPy8EG_A14N23LO_NUHM2_m12_350_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394309.MGPy8EG_A14N23LO_NUHM2_m12_400_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394310.MGPy8EG_A14N23LO_NUHM2_m12_400_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394311.MGPy8EG_A14N23LO_NUHM2_m12_400_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394312.MGPy8EG_A14N23LO_NUHM2_m12_400_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394313.MGPy8EG_A14N23LO_NUHM2_m12_500_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394313.MGPy8EG_A14N23LO_NUHM2_m12_500_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6241 mc15_13TeV.394314.MGPy8EG_A14N23LO_NUHM2_m12_500_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394315.MGPy8EG_A14N23LO_NUHM2_m12_500_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394316.MGPy8EG_A14N23LO_NUHM2_m12_500_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394317.MGPy8EG_A14N23LO_NUHM2_m12_600_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394317.MGPy8EG_A14N23LO_NUHM2_m12_600_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6241 mc15_13TeV.394318.MGPy8EG_A14N23LO_NUHM2_m12_600_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394319.MGPy8EG_A14N23LO_NUHM2_m12_600_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394320.MGPy8EG_A14N23LO_NUHM2_m12_600_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394321.MGPy8EG_A14N23LO_NUHM2_m12_700_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394322.MGPy8EG_A14N23LO_NUHM2_m12_700_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394323.MGPy8EG_A14N23LO_NUHM2_m12_700_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394324.MGPy8EG_A14N23LO_NUHM2_m12_700_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394325.MGPy8EG_A14N23LO_NUHM2_m12_800_weak_C1C1_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394326.MGPy8EG_A14N23LO_NUHM2_m12_800_weak_N2C1m_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394327.MGPy8EG_A14N23LO_NUHM2_m12_800_weak_N2C1p_2LMET50_MadSpin.evgen.EVNT.e6233 mc15_13TeV.394328.MGPy8EG_A14N23LO_NUHM2_m12_800_weak_N2N1_2LMET50_MadSpin.evgen.EVNT.e6233
Use ami show dataset info: ami show dataset info mc15_13TeV.*.MGPy8EG_A14N23LO_NUHM2_m12_*_weak_*_2LMET50_MadSpin.evgen.EVNT.e6233
Results:

m12	crossSection_mean	GenFiltEff_mean
300_C1C1	9.5000E-05	1.4336E-01
300_N2C1m	1.3317E-05	2.5442E-01
300_N2C1p	2.3558E-05	2.5151E-01
300_N2N1	3.4970E-13	3.0253E-01
350_C1C1	7.6915E-05	1.2992E-01
350_N2C1m	5.1398E-05	2.5287E-01
350_N2C1p	8.9778E-05	2.5578E-01
350_N2N1	1.3010E-13	2.2768E-01
400_C1C1	6.7013E-05	1.2201E-01
400_N2C1m	5.2073E-05	2.2064E-01
400_N2C1p	9.1605E-05	2.2389E-01
400_N2N1	8.8153E-14	2.0416E-01
500_C1C1	5.7197E-05	1.0888E-01
500_C1C1 (e6241)	5.7163E-05	1.0822E-01
500_N2C1m	5.2888E-05	1.8924E-01
500_N2C1p	9.3905E-05	1.8806E-01
500_N2N1	3.9978E-14	1.7602E-01
600_C1C1	5.2260E-05	9.8353E-02
600_C1C1 (e6241)	5.2268E-05	1.0002E-01
600_N2C1m	5.3768E-05	1.6871E-01
600_N2C1p	9.5337E-05	1.6926E-01
600_N2N1	1.9708E-14	1.5353E-01
700_C1C1	4.9200E-05	9.3538E-02
700_N2C1m	5.4490E-05	1.5801E-01
700_N2C1p	9.6593E-05	1.5874E-01
700_N2N1	1.0495E-14	1.3788E-01
800_C1C1	4.7195E-05	8.7180E-02
800_N2C1m	5.5645E-05	1.3915E-01
800_N2C1p	9.7683E-05	1.4631E-01
800_N2N1	6.0353E-15	1.2825E-01

Download EVNT --> Done!
ouhep04@cern: /UserDisk2/yushen/NUHM2/evgen/
ouhep05@nhn: /raid05/atlas/data/NUHM2/evgen/
Download SUSY16 DAOD --> Done!
ouhep04@cern: /UserDisk2/yushen/NUHM2/SUSY_DAOD/p2952/
ouhep05@nhn: /raid05/atlas/data/NUHM2/SUSY_DAOD/p2952/
Download TRUTH3 DAOD --> Done!
ouhep04@cern: /UserDisk2/yushen/NUHM2/TRUTH3_DAOD/
ouhep05@nhn: /raid05/atlas/data/NUHM2/TRUTH3_DAOD/
Use new TRUTH3 samples to run SimpleAnalysis
Download Joey's ntuple v1.9 for NUHM2 --> Done!
ouhep04@cern: /afs/cern.ch/user/y/yushen/afsWorkingArea/private/Higgsino/MC/
ouhep05@nhn: /raid05/atlas/data/Higgsino/
Update my cutflow framework to run Joey's new ntuple.
Run my cutflow framework and make the significance plot.
Read the internal note and make modification if needed.

ytatus94 commented 6 years ago

20171004

Judita Mamuzic: (04.10.17 3:17 AM)
/afs/cern.ch/work/j/jeliu/public/HiggsinoFitterTrees/v1_9/

bugs = ['other',
'diboson',
'Zttjets',
'top',
'fakes',
'data'
]

/raid05/users/mamuzic/public/ForChris_MCProd/
https://svnweb.cern.ch/trac/atlasphys-susp/brower/Physics/SUSY/Papers/Electroweak2017/Compressed/paper
--doNewFFs

HistFitter.py -t -w -f -D "corrMatrix" -F excl -g 'MGPy8EG_A14N23LO_NUHM2_m12_300_weak_2LMET50_MadSpin'
$HISTFITTER/analysis/higgsino/HiggsinoConfig.py -u " --AllSys --doNewFFs" && HistFitter.py -F excl -p -g 'MGPy8EG_A14N23LO_NUHM2_m12_300_weak_2LMET50_MadSpin'
$HISTFITTER/analysis/higgsino/HiggsinoConfig.py -u " --AllSys --doNewFFs" && HistFitter.py -F excl -l -q 'MGPy8EG_A14N23LO_NUHM2_m12_300_weak_2LMET50_MadSpin'
$HISTFITTER/analysis/higgsino/HiggsinoConfig.py -u " --AllSys --doNewFFs" &&

ytatus94 commented 6 years ago

2017/10/17
- SimpleAnalysis:
- If I fill histograms directly in the SimpleAnalysis code itself, the mcEventWeight will automatically be used.
- If I store ntuple and want to fill the histogram, then I have to Fill the weight.
- I don't need to do float mc_weight = event->getMCWeights()[0]; and fill it in the ntuple. It is saved in eventWeight variable automatically.

ytatus94 commented 6 years ago

2017/10/19
- HistFitter
```
setupATLAS
setup root
source setup.sh
-h help
-t  create histograms: if the cut, weight, systematics, then the histograms need to be re-generated.
-w create workspace
-f  fit the workspace
-l  limit plot
-D  draw plot: -D "[before/after/all plot/corrMatrix/setcomponents/likehood]"
-d  draw plot
-w -f -i
```
- make_table.py, YieldsTable.py, SysTable.py
- hypostest_to_cvs
- upper limit.root
- make list files
- Run --NoSys first
- exclusion fit for NUHM2
- Christopher Arnold Walker: (04:42 PM) Hi Yu-Ting, here is the link to where I've been running histfitter: /raid05/users/walker/judita/ForChris_MCProd/HiggsinoFitter_mcprod I've been using the script "spawnthem.sh" I run in the run/ subdirectory with HiggsinoConfig.py located in the same folder, after copying the directory run the first_setup.sh script. run the setup in the same directory as the script, but do the usual setupATLAS first you cad delete the extra run.../ subdirectories first in the run/ subdirectory you can delete the extra subdirectories. They will be recreated when you run

ytatus94 / Higgsino

My logs #3

N1	N2	N3	N4	C1	C2
N1	CC	CC	CA	CA	CC	CA
N2	CC	CC	CA	CA	CC	CA
N3	AC	AC	AA	AA	AC	AA
N4	AC	AC	AA	AA	AC	AA
C1	CC	CC	CA	CA	CC	CA
C2	AC	AC	AA	AA	AC	AA

m12	N2N1	C1C1	N2C1p	N2C1m
300	117	123	146	129
350	197	137	129	139
400	221	135	197	172
500	266	171	234	200
600	281	160	249	246
700	303	185	234	229
800	298	180	275	233

N1	N2	N3	N4	C1	C2
N1	CC	CC	CA	CA	CC	CA
N2	CC	CC	CA	CA	CC	CA
N3	AC	AC	AA	AA	AC	AA
N4	AC	AC	AA	AA	AC	AA
C1	CC	CC	CA	CA	CC	CA
C2	AC	AC	AA	AA	AC	AA

m12	N2N1	C1C1	N2C1p	N2C1m
300	117	123	146	129
350	197	137	129	139
400	221	135	197	172
500	266	171	234	200
600	281	160	249	246
700	303	185	234	229
800	298	180	275	233

N1	N2	N3	N4	C1	C2
N1	CC	CC	CA	CA	CC	CA
N2	CC	CC	CA	CA	CC	CA
N3	AC	AC	AA	AA	AC	AA
N4	AC	AC	AA	AA	AC	AA
C1	CC	CC	CA	CA	CC	CA
C2	AC	AC	AA	AA	AC	AA

m12	N2N1	C1C1	N2C1p	N2C1m
300	117	123	146	129
350	197	137	129	139
400	221	135	197	172
500	266	171	234	200
600	281	160	249	246
700	303	185	234	229
800	298	180	275	233