Some doubts on Contour Plots

ROOT
,
1

Please read tips for efficient and successful posting and posting code

ROOT Version: 6.14/04
Platform: linuxx8664 (Ubuntu 16.04.12)
Compiler: gcc 5.4.0

Hi,
I have a few questions on Contour plots. Basically, I am learning to make contour plots.

  1. I was trying to make a contour plot from a Delphes output ROOT file. I wanted the as Z-axis the event numbers or the constituent particle number of that branch. I am not sure which variable it’s taking as the Z-axis, but it cannot be the number of events. How to set Z-axis as the number of events. I have attached a Sample macro and a Dropbox link to the ROOT files. Also, I am attaching the contour that I could make.
    SampleMacro.C (1.6 KB)
    ROOT Files

    Photon%20Contour
    Photon Contour.png1108×471 65.2 KB

  2. I have an ASCII file with three columns. The first two columns are two variables and the third column is the function of them. I want to plot a contour of the two variables with the 3rd column as the Z-axis. How to do that? I am also attaching a small sample data file.
    data.txt (2.8 KB)

Regards,
Saumyen

2

The “Z-axis” is the number of entries in histograms’ bins (seems that bins in your histograms have between 0 and 13 entries).

3

Thanks, Wile.
Yes, that I noticed. The thing is that when I was plotting Eta and Phi 1-D histograms, they were okay as you already know. But when I am trying to plot contour why the z-axis is taking such a range of values? In the input ROOT files there are more than 9k events for Photon.
Do have any idea?

Regards,
Saumyen

4

Your “signalSP.root” has 10000 entries in the “Delphes” tree.
You create a 50 bins x 50 bins histogram which means that in average you can expect 4 entries per bin.

5 
{
  TGraph2D *g = new TGraph2D("data.txt");
  g->SetMarkerStyle(20);
  g->Draw("pcolz"); // "pcolz" or "colz" or "surf2z" or ...
  gPad->SetLogz(1);
}
6

Thank you so much, Wile. This is working fine. Just one question: if I have more than 3 columns then should I do the same thing as in TGraph1D to read the columns, i.e., TGraph2D *g = new TGraph2D("data.txt", "%*lg %lg %lg %lg %*lg"); or anything else?

Regards,
Saumyen

7

https://root.cern/doc/master/classTGraph2D.html#a43266b55df2d9c83fee12f587f405b25

8

Thanks a lot Olivier. Sorry, I should have noticed that.
Thanks again.

Regards,
Saumyen

9

Hi Wile and Olivier,
Sorry to again bother you on the same topic. When I tried the above commands as a macro it worked fine with a data file with only 3 columns. But when I tried running the same macro with a data file with multiple columns and selecting specific columns it gives error messages.

Error in <TGraph2D::Paint>: Empty TGraph2D

I am attaching both the data file and the macro. Can you tell me what’s wrong here?
data.txt (2.8 KB)
not_working.txt (247.2 KB)
Test.C (199 Bytes)

NOTE that when I’m working with the file ‘working.txt’, I am deleting the first %*lg in line-2.

Regards,
Saumyen

10

Okay, sorry. It’s now working when I replace the %*lg with %*s.
But how do I interchange the X-axis and Y-axis?

Regards,
Saumyen

11

If you do not want to change your macro you will need to invert the columns in data.txt:

awk '{ t=$1 ; $1=$2; $2=t; print }' data.txt > data2.txt
12
13

Thanks, Olivier and Wile. Both the suggestions working like charm.
In particular, I liked command suggested by Olivier. It’ll save me a lot of time in other cases also. But also, I noticed that in the new file it’s replacing the column separator from ‘tab’ to ‘space’. Is that expected?
Anyway, thanks a lot.

Regards,
Saumyen

14

If your file has three columns, and you want to invert the first two, and separate them with tabs do:

 awk '{ print $2"\t"$1"\t"$3 }' data.txt > data2.txt

There is many awk tutorials on the web.

15

Thanks. This is working fine…
Thanks a lot Olivier.

Regards,
Saumyen

16

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.