giovedì 27 settembre 2007

sample in sotto-regioni

da Mattia:

ho caricato su castor anche i dati con i bordi sovrapposti, esattamente in $CASTOR_HOME/ECALCalib/AlCa/RelVal/overlap/
quelli con i bordi affiancati sono su $CASTOR_HOME/ECALCalib/AlCa/RelVal/disjoint/

i numeri 1, 4, 6 hanno il solito vecchio problema del pt()!=0, non so come isolare gli eventi corrotti

Castor files

Stored some data in Castor

at $CASTOR_HOME/EcaLCalib/AlCa/RelVal/disjoint/ there are 9 files (EBFrontier*.cfg) which scan Barrel in disjoint zones (crystals with ieta 0-10, 10-20, etc)

at $CASTOR_HOME/EcaLCalib/AlCa/RelVal/overlap/ there are 9 files (EBFrontierZoneSovrapposte*.cfg) which scan Barrel in overlap zones (crystals with ieta -5 to 15, 5 to 25, etc)

lunedì 24 settembre 2007

Frontier filtered files

A small comunication after a long run. We got new data in the 700Mb file


produced from Frontier events filtered between etamin 0.1 and etamax 0.3.

EDIT: thanks to cmsmi3 nightwork we got the same for Endcap,


filtered by etamin = 1.6 and etamax = 1.8

25 september, 14:15 update :

data on IV module:

a new little tool

We got a new tool, in file

It compares two sets of calibration coefficients and produces the distribution of the ratios.
Here is an example of the output:

Of course it works on Barrel and both Endcaps.

giovedì 20 settembre 2007

Again on Neighbourhoods

Today I've commited two new graphs, which represents the distribution of energy around the most energetic crystals.

In the first one (of whom I post a view in 2d and one in 3d) we can see Energy in the 5x5 neighbourhood. It's been runned on all events, assigning (0,0) coordinates to the crystal with maximum Energy.

In the latter graph there is the rate of the Energy in the central crystal on the total Energy in the neighbourhood.


mercoledì 19 settembre 2007

Neighbourhoods Energy

New methods Energy25Barrel and Energy25Endcap give back the total energy of the 5x5 neighbourhood around the crystal with max energy. In input they need the coordinates (eta, phi) or (x,y,z) where z identifies which endcap is, and the rechitcollection.

We can use them in order to recalculate EoP.

In the next two graphs here's the results, runned on


In blue "classic" EoP, in red EoP with E re-evaluated on the 5x5 neighbourhood.

martedì 18 settembre 2007

Eta filtered files

The first etafiltered file is now in /data/calib/csa07/alcaReco/EB100pb.root
it has most of the electrons between ieta 1 and ieta 15 in CMSSW reference system.
Most of the events are also between iphi20 and iphi60
It comes from the 100pb set of files.
A small selection only for the endcap is being processed right now.
It will be called EE100pb.root and stored in the same place.
This files will also be on my castor to be accesible for running from the lxplus and from the lsf.
They are in /castor/

lunedì 17 settembre 2007

100pb file

We processed all the 100pb file and here is a glance at how they look like.
You can see the Energy distribution, the number of the electrons in a xtal and the EoP distribution.
Good night!

domenica 16 settembre 2007

Sunday's news

my first post here!

committed alcaelectronstest.h and AlCaElectronsTest_filtered.root
I've tried to run with 100 events, every TH1F and TH2f looks to work properly.

we got

OBJ: TH1F pDistro_endcap pDistro_endcap : 0 at: 0x9421028
KEY: TH2F m_barrelGlobalCrystalsMap;1 m_barrelGlobalCrystalsMap
KEY: TH2F m_barrelLocalCrystalsMap;1 m_barrelLocalCrystalsMap
KEY: TH2F m_endcapGlobalCrystalsMap;1 m_endcapGlobalCrystalsMap
KEY: TH2F m_endcapLocalCrystalsMap;1 m_endcapLocalCrystalsMap
KEY: TH2F m_barrelGlobalCrystalsEnergy;1 m_barrelGlobalCrystalsEnergy
KEY: TH2F m_barrelLocalCrystalsEnergy;1 m_barrelLocalCrystalsEnergy
KEY: TH2F m_endcapGlobalCrystalsEnergy;1 m_endcapGlobalCrystalsEnergy
KEY: TH2F m_endcapLocalCrystalsEnergy;1 m_endcapLocalCrystalsEnergy
KEY: TH2F m_barrelGlobalCrystalsEnergyMap;1 m_barrelGlobalCrystalsEnergyMap
KEY: TH2F m_endcapGlobalCrystalsEnergyMap;1 m_endcapGlobalCrystalsEnergyMap
KEY: TProfile2D m_barrelGlobalCrystalsMomentum;1 m_barrelGlobalCrystalsMomentum
KEY: TH1F pDistro_barrel;1 pDistro_barrel
KEY: TH1F Edistro_barrel;1 Edistro_barrel
KEY: TH1F EoP_barrel;1 EoP_barrel
KEY: TProfile2D m_endcapGlobalCrystalsMomentum;1 m_endcapGlobalCrystalsMomentum
KEY: TH1F pDistro_endcap;1 pDistro_endcap
KEY: TH1F Edistro_endcap;1 Edistro_endcap
KEY: TH1F EoP_endcap;1 EoP_endcap

as a example:

sabato 15 settembre 2007

Saturday's news

The 100pb have been processed and put where i told before on castor. They won't be on cmsmi3 till this evening.
422Kevents succeded while ~129K still needs to be reprocessed...Hoping in castor.

Data processing

The 10pb data have almost all been processed. Some are still not available on castor but we have 422Kevents. The are in /castor/
They are currently been transferred also in cmsmi3: /data/calib/csa07/alcaReco/10pb/
Each file is about 7MB.
We'll try to recover another ~100Kevents from the processes aborted because castor didn't feed the data fast enough for the LSF.

Now the batch system is processing, actually really fastly, the 100 pb datas.
They will be on /castor/ Then we'll transfer that on /data/calib/csa07/alcaReco/100pb/

As now we have 130Kevents processed, and they have been processed just in 30 minutes.
Tomorrow we'll process the frontier.

Then we'll run the alcaElectron test on all the data to see how does the data look like and well start deep testing the InvMatrixLooper.


mercoledì 12 settembre 2007

The draw of a Region

Now we tested the region Definition.
In this plot we show with different color even and odd regions.
The number are the id of the xtals in that specific region

CalibBlock's Test

This morning the EcalCalibBlock class underwent some test.

We filled a map of 25 entries with random number that acted like the "energy deposit"
We summed all the entries and this was the "momentum".
With this values we filled the calibBlocks and we repeated the procedure for many times. (with only one filling the system isn't determined)
That we invoked the solve method and then read back the coefficients. If the class is working properly the results should all be one.
And that's what happened.
Then we multiplied the entries of the map for some random coefficients, after summing this numbers for the "momentum". If the class was working the results should have been the inverse of our input coeffs. And that's what happened.
The program that does this test is in EcalAlCaRecoProducers/bin/

lunedì 3 settembre 2007

Ring definitions

Now we are ready.
Finally we solved all the majors errors in the InvRingCalib code about the first part that is about getting the geometry from the event setup and use it to define what is a ring in the Ecal Endcap.
Here are some graphs, as always we apologize for the horrible colours that root gives us.
In the cfg we specified all the rings (from 0 to 39) with regions 4 ring wide.
In the barrel eta start 3, eta end 159, 4 rings wide.
In the endcap 0(green) is where there aren't xtals, 10 in some regions and -10 in the following ones.