[[root:root|Getting started with ROOT]] -> [[root:pyroot|Useful pyROOT snippets]] → [[root:pyroot_th1|Handling TH1 histograms]]

====== Handling TH1 histograms ======

[[https://root.cern.ch/doc/master/classTH1.html|Here]] is the class reference for ''TH1''.


===== Importing ROOT classes =====

To use ''TH1D'' in ''python'', you can either only import a ''TH1D'' class:
<code python>
from ROOT import TH1D
hist = TH1D("hist_name","hist_title",100,0,100)
</code>
or //all// classes:
<code python>
from ROOT import * # loads all classes, and thus takes longer
hist = TH1D("hist_name","hist_title",100,0,100)
</code>
or only the ''ROOT'' module:
<code python>
import ROOT
hist = ROOT.TH1D("hist_name","hist_title",100,0,100)
</code>
The examples below also make use of ''ROOT'''s ''gRandom'' to make some random variables.
===== Filling a histogram =====

Simple enough:

<code python>
hist = TH1D("hist_name","hist title",100,0,100)
for i in xrange(10000):
    hist.Fill(gRandom.Gauss())
</code>

With some weight:
<code python>
weight = gRandom.Gauss(1,0.1)
hist.Fill(gRandom.Gauss(),weight)
</code>

===== Filling a histogram from a tree =====

The classic way would be with a simple for loop
<code python>
hist = TH1D("hist_name","hist title",100,0,100)
for event in tree:
    hist.Fill(tree.px)
</code>

But the simplest and most powerful way to draw some variable into a histogram, is with the [[https://root.cern.ch/doc/master/classTTree.html#a73450649dc6e54b5b94516c468523e45|TTree::Draw function]]:
<code python>
hist = TH1D("hist_name","hist title",100,0,100)
tree.Draw("px >> hist_name")
</code>
Note that the histogram has been saved into ''ROOT'''s working memory and is therefore accessible via its name in the ''Draw'' function, ''"hist_name"'' in this example.

You can also use **mathematical expressions**, including **boolean expressions** (which can take the value ''0.0'' or ''1.0''):
<code python>
tree.Draw("2*sqrt(px*px+py*py) >> hist_name")
tree.Draw("abs(eta) >> hist_name")
tree.Draw("pt1*(eta1<2.4) + pt2*(eta2<2.4) >> hist_name") # boolean expressions
</code>
This also includes custom functions, which you can load into ''ROOT'' with a ''C'' file:
<code python>
from ROOT import gROOT, ...
gROOT.Macro("myfunctions.C+")
...
tree.Draw("deltaR(eta1,phi1,eta2,phi2) >> hist_name")    # deltaR     function defined in "myfunctions.C+"
tree.Draw("pt1 >> hist_name","(pt>30)*weightEta1(eta1)") # weightEta1 function defined in "myfunctions.C+"
</code>

You can apply some **selections** using variable, available in the tree's branches:
<code python>
tree.Draw("px >> hist_name","pt>20 && E>20")
</code>

You can apply some selections and/or a **weight**, if a weight variable is available in the tree:
<code python>
tree.Draw("px >> hist_name","weight")
tree.Draw("px >> hist_name","(pt>20 && E>20)*weight")
</code>

An equivalent method is creating the histogram in the ''Draw'' method. If you need the ''TH1F'' object, you can still get it with ''gDirectory'' (see [[root:pyroot_TFile#gDirectory|here]]):
<code python>
from ROOT import gDirectory
...
tree.Draw("pt >> h(100,0,100)")
hist = gDirectory.Get("h") # will be a TH1F object
</code>

It's also possible to pass some **draw option** you would normally use in "TH1::Draw()". It is passed as the third string.
<code python>
tree.Draw("px >> h1(100,0,100)","","E2")
tree.Draw("py >> h2(100,0,100)","","E2 SAME")

tree.Draw("px >> h1(100,0,100)","abs(eta)>2","E2")
tree.Draw("px >> h2(100,0,100)","abs(eta)<2","E2 SAME")
</code>

__Protip__: This is very useful for quick drawing and comparing, for example in the ''ROOT'' command line.
<code C++>
$ root -l tree.root
[1] tree->Draw("px >> h1(100,0,100)","")
[2] tree->Draw("px >> h1(100,0,100)","pt>20")
[3] tree->Draw("px >> h1(100,0,100)","","E2")
[4] tree->Draw("py >> h2(100,0,100)","","E2 SAME") // compare to px
</code>