How to simply make a histogram from .dat file

[cocoleia]

---

_ROOT Version:ROOT 6.12/06
_Platform:Debian?
_Compiler:

---

I have a .dat file, it is separated in 2 columns and about 20 000 rows.
I want to open this file somehow and plot a histogram for one of the columns.

I have looked at these two threads:

How can I make a histogram with a .csv file?

Hello, I’m new to ROOT, and was hoping that someone could help me out. My goal is to be able to create a histogram from data stored in a .csv file. But I don’t know how to convert a .csv file to a .txt or .dat file. Does anyone know how to do this using ROOT? Thanks!

Making a .csv file into a histogram

Hello, I have the following problem. I convert the csv files into ROOT histograms using the following algorithm: TNtuple *t=new TNtuple("Data","title","X:Y"); t->ReadFile(a); t->Draw("X","Y"); Then I save the histogram in .ps or .root format. But the binning seems to change in the process (the fluctuations are considerably larger; for comparison, I provide one of my ROOT histograms and the same histogram, done in Excel). Is there any way to change the binning? Or maybe there’s a better wa…

and tried this code but it does not work for me.

I understand how to draw a histogram based off of things that I have calculated (example: write a code to calculate the energy then make a histogram of this energy)

I just don’t know what command to use to read my .dat file and make the histogram

I am completely new to c++ and root

[Danilo]

Hi,

can you attach the first 10 or 20 lines of the file or share it entirely so that I can produce an example which works?

Best,
D

[cocoleia]

So here there is a number between 1 and 5, and then another value next to it.

Right now I want to plot the other value, regardless of whether it is 1,2,3,4,5

I converted to a .txt file and attached. I assume the code is the same for .txt and .dat? It won’t let me attach .datevents.txt (292.2 KB)

[behrenhoff]

The code shown above works just fine with your file.

So what do you mean exactly with “it does not work for me.”?

[cocoleia]

Which code then? For me, root just kept “crashing” and nothing would run.

I also want to know how to choose certain elements to plot.

Let’s say I would want to plot only the numbers that are with 1, how would I do it ?

[cocoleia]

If you’re talking about this
auto tree = TTree::ReadFile(events.dat,“x:y:z”);
tree->Draw(“x”);
I guess I don’t know how to write the file name in a way that works, since it doesn’t accept it like this. Also don’t know what “draw(x)” will be drawing (first column, etc)

or if I need to change x,y,z

[Wile_E_Coyote]

TTree *t = new TTree("t", "my events");
t->ReadFile("events.dat", "first/I:second/D");
t->Print();
t->Draw("first");

[cocoleia]

@Wile_E_Coyote

TTree *t = new TTree(“t”, “my events”); t->ReadFile(“events.dat”, “first/D:second/D”); t->Draw(“first”);

ah, I see. Thank you.

I have another question, if I wanted to select to plot only the energies corresponding to 1, or to 2 etc how would I arrange this ?

[Wile_E_Coyote]

t->Draw("second", "first == 1");

[Danilo]

Hi,

this is the modern way to achieve this:

auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("data.txt", false, ' ');
auto h = df.Histo1D("Col1")
h->Draw();

It has plenty of advantages, besides being the suggested way to code:

1. You can easily convert your dataset to ROOT:

auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("data.txt", false, ' ');
df.Snapshot("data", "data.root"); // <- one line is enough
auto h = df.Histo1D("Col1")
h->Draw();

2. You can run using all the cores of your machine (not mutually exclusive with 1.)

ROOT::EnableImplicitMT(); // <- just adding this line is enough
auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("data.txt", false, ' ');
df.Snapshot("data", "data.root");
auto h = df.Histo1D("Col1")
h->Draw();

3. You can create several histograms in the same event loop, therewith limiting the disk access:

ROOT::EnableImplicitMT(); // <- just adding this line is enough
auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("data.txt", false, ' ');
df.Snapshot("data", "data.root");
auto h = df.Histo1D("Col1");
auto h1 = df.Histo1D("Col0");
auto hWeighted = df.Histo1D("Col1", "Col0");
auto h2 = df.Histo2D("Col0", "Col1");
// ...

See all different options in these examples.

Best,
D

[cocoleia]

@Wile_E_Coyote

@Danilo

Hi,

this is the modern way to achieve this:

auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("data.txt", false, ' ');
auto h = df.Histo1D("Col1")
h->Draw();

These are both very helpful.
One last thing, if I wanted to make a 2D TH2F distribution, how would I proceed ?

[Danilo]

Hi,

it’s in my point 2):

auto h2 = df.Histo2D("Col0", "Col1");

Cheers,
D

[cocoleia]

@Danilo

Hi,

it’s in my point 2):

auto h2 = df.Histo2D("Col0", "Col1");

When I run it for the 1D, I simply get a blank canvas. When I run with this line for the 2D, I get many errors

error: no matching member function for call to ‘Histo2D’

among other things.

Sorry if these questions are basic, I am very new to this

[Danilo]

Hi,

these lines have been tested on my machine and I am sure they work. I think we are missing something fundamental Can you share in a post the code which leads to the errors you are reporting? I am sure we can solve those quickly.

Cheers,
D

[cocoleia]

void two_d()
{
	auto df = ROOT::Experimental::TDF::MakeCsvDataFrame("events.dat", false, ' ');
	auto h2 = df.Histo2D("Col0", "Col1");
	h2->Draw();
    
}

I really don’t know what to add or to write

[Danilo]

Hi,

can you try:

      auto df = ROOT::Experimental::TDF::MakeCsvDataFrame( "data.txt", false, ' ');
      auto h = df.Histo2D({"myHisto", "My Title", 5, 0.5, 5.5, 256, -400, 400}, "Col0", "Col1");
      h->Draw();

Cheers,
Danilo

[cocoleia]

@Danilo

auto df = ROOT::Experimental::TDF::MakeCsvDataFrame( “data.txt”, false, ’ '); auto h = df.Histo2D({“myHisto”, “My Title”, 5, 0.5, 5.5, 256, -400, 400}, “Col0”, “Col1”); h->Draw();

This is what I get

Processing Histo2D.C…
terminate called after throwing an instance of ‘std::runtime_error’
what(): Unknown columns: Col0,Col1

[cocoleia]

@Danilo

auto df = ROOT::Experimental::TDF::MakeCsvDataFrame( “data.txt”, false, ’ '); auto h = df.Histo2D({“myHisto”, “My Title”, 5, 0.5, 5.5, 256, -400, 400}, “Col0”, “Col1”); h->Draw();

Now it runs, but it only produces a blank canvas c1

[Danilo]

Ok.
This is a well known behaviour independent of data frame. You need to make sure the object survives after the end of the scope. For example:

     auto df = ROOT::Experimental::TDF::MakeCsvDataFrame( "data.txt", false, ' ');
      auto h = df.Histo2D({"myHisto", "My Title", 5, 0.5, 5.5, 256, -400, 400}, "Col0", "Col1");
      auto mycanvas = new TCanvas();
      h->DrawCopy();

[cocoleia]

This works, but it produces a result not like I had expected. Glad to see it working though. Thanks for your help !