Adding new Branches to existing TTree or TNtuple

ROOT
1

I have simple code to write new branches to an existing TTree, but the resulting root file seems to contain two copies of the same ttree. Can anyone see why in the code below?

The problem I’m trying to solve is that I have an existing TTree that contains a few arrays of known length. I’m writing a function to loop through all entries in the TTree, and then loop over all elements in the array to find elements that pass certain cuts and place those in a new branch.

I’m still thinking how I can simplify the code below, or if it could be made faster.

Code here:

[code]from ROOT import TFile, TTree # Import any ROOT class you want
from array import array # used to make Float_t array ROOT wants
import sys

ttreeName = “NTuples/Analysis” # TTree name in all files

listOfFiles = [“testing.root”]

for fileName in listOfFiles:
file = TFile(fileName, “update”) # Open TFile
if file.IsZombie():
print “Error opening %s, exiting…” % fileName
sys.exit(0)
print “Opened %s, looking for %s…” % (fileName, ttreeName)

ttree = TTree()                           # Create empty TTree, and
try:                                      # try to get TTree from file.
    file.GetObject(ttreeName, ttree) 
except:
    print "Error: %s not found in %s, exiting..." % (ttreeName, fileName)
    sys.exit(0)
print "found."

# Add those variables into the TTree
print "Adding new branches:\n  ",
listOfNewBranches = []
newJetPt   = array( 'f', [0] )
listOfNewBranches.append( ttree.Branch("passjetPt", newJetPt, "passjetPt/F") )
newJetEta  = array( 'f', [0] )
listOfNewBranches.append( ttree.Branch("passjetEta", newJetEta, "passjetEta/F") )
newJetPhi  = array( 'f', [0] )
listOfNewBranches.append( ttree.Branch("passjetPhi", newJetPhi, "passjetPhi/F") )
newJetEmEnergyFraction = array( 'f', [0] )
listOfNewBranches.append( ttree.Branch("passjetEmEnergyFraction", newJetEmEnergyFraction, "passjetEmEnergyFraction/F") )
newJetFHPD = array( 'f', [0] )
listOfNewBranches.append( ttree.Branch("passjetFHPD", newJetFHPD, "passjetFHPD/F") )

# Loop over all the entries
numOfEvents = ttree.GetEntries()
for n in xrange(numOfEvents):
    newJetPt[0]  = 0.0
    newJetEta[0] = 0.0
    newJetPhi[0] = 0.0
    newJetEmEnergyFraction[0] = 0.0
    newJetFHPD[0] = 0.0
    ttree.GetEntry(n)
    for i in 0,1,2,3:      # Loop over the top 3 jets until we find one passing cuts
        if ttree.jetPt[i] < 5.0: break
        if (ttree.emEnergyFraction[i]>0.01) and (ttree.fHPD[i]<0.98):
            # Found a jet that passes cuts
            newJetPt[0]   = ttree.jetPt[i]
            newJetEta[0]  = ttree.jetEta[i]
            newJetPhi[0]  = ttree.jetPhi[i]
            newJetEmEnergyFraction[0] = ttree.emEnergyFraction[i]
            newJetFHPD[0] = ttree.fHPD[i]
            break
    # Fill new branches
    for newBranch in sorted(listOfNewBranches):
        newBranch.Fill()

file.Write()
file.Close()[/code]
2

[quote]but the resulting root file seems to contain two copies of the same ttree.[/quote]How so? Are you sure they are not simply 2 revisions of the TTree object (i.e. just the meta data; see the User’s Guide for more information about object revision in a TFile).

Cheers,
Philippe.

3

Can I ask which chapter in the user’s guide you’re referring to?
root.cern.ch/drupal/content/users-guide

I only thought it was two ttrees because I saw the tree (named “Analysis”) listed twice in the TBrowser window. But I see now the file size did not double.

4

Hi,

I am refering to the Input/Output chapter. The concept I am speaking of is named ‘cycle’. In your case, you definitively just have 2 cycles for the TTree which is normal and expected (and does not mean that the data of the TTree has been duplicated, just that 2 revisions of the TTree meta data is available (one slightly older than the other)).

Cheers,
Philippe.