Text labels won't automatically rotate to fit

David_Vannerom · May 27, 2022, 8:41am

Dear ROOTers,

I’m having a problem with user-defined axis labels. I’d like the X-axis labels to rotate from a certain angle in order not to overlap onto each other. I used to have it work without the ratio plot (it was done automatically), but now that I’ve added this extra pad, they stay horizontal for some reason. Any idea why and how to get back to the default, rotating behaviour?

I’m uploading the macro (it’s in pyROOT actually) here:

makeSelectionEfficiency.py (14.2 KB)

It should be ran as:

python makeSelectionEfficiency.py -m 100 -q 1

You will also need the 4 root files (M100_Q1_*.root) found at:

https://t2bweb.iihe.ac.be/~vannerom/FCP/MuonHack/

I also attach the output plot in pdf.

selectionEfficiency_M100_Q1.pdf (15.8 KB)
Thanks!

ROOT Version: 6.24/06
Built for linuxx8664gcc on Sep 02 2021, 14:20:23
From tags/v6-24-06@v6-24-06

dastudillo · May 27, 2022, 9:50am

The macro is missing.

David_Vannerom · May 27, 2022, 9:55am

Thanks! Sorry about that, it’s in the edited post now.

David_Vannerom · May 30, 2022, 12:35pm

Hello,

Any idea on this?

Thanks!

couet · May 30, 2022, 12:56pm

It works automatically on for the alpha-numeric histograms.

David_Vannerom · May 30, 2022, 3:03pm

What does that mean? It used to work with the exact same plot and labels but just without the ratio plot so I’m confused.

couet · May 30, 2022, 3:40pm

Ah ok. I’ll have closer look.

couet · May 30, 2022, 3:44pm

I removed the PATH in your script but still it does not work.
Do you have a running version without doing too much editing ?

 % python3 makeSelectionEfficiency.py -m 100 -q 1
Filling file M100_Q1_analysisSample.root
Error in <TFile::TFile>: file /Users/couet/Downloads/M100_Q1_analysisSample.root does not exist
Traceback (most recent call last):
  File "makeSelectionEfficiency.py", line 341, in <module>
    fillEfficiency(h_noHack_passed,h_noHack_total,nBins,mass,charge,0)
  File "makeSelectionEfficiency.py", line 16, in fillEfficiency
    f = ROOT.TFile(fileName)
  File "/Users/couet/git/couet-root-bin/lib/ROOT/_pythonization/_tfile.py", line 90, in _TFileConstructor
    raise OSError('Failed to open file {}'.format(args[0]))
OSError: Failed to open file M100_Q1_analysisSample.root

David_Vannerom · May 30, 2022, 5:27pm

Could you try to have the script and the 4 root files in the same folder and change:

fileName = "/user/vannerom/FCP/BatchArea/CMSSW_10_2_3/src/AnalyzerArea/FCPTreeMaker/M"+mass+"_Q"+charge+"_"

to

fileName = "M"+mass+"_Q"+charge+"_"

?

Let me know how it goes, thank you.

dastudillo · May 31, 2022, 5:25am

For me it works as expected, on WSL Ubuntu 20.04, python 3.8.10, ROOT 6.26/02

David_Vannerom · May 31, 2022, 10:12am

That is weird. I tried it locally on my computer (python 3.8.10, ROOT 6.24/06) and it still doesn’t work. But I don’t expect this to be a version issue anyway since it used to work without the ratio plot.

David_Vannerom · June 1, 2022, 8:50am

@couet were you able to run this? If yes, did it also work for you or do you see what I see, i.e. overlapping labels? Thanks

couet · June 1, 2022, 9:18am

I still get:

Error in <TFile::TFile>: file /Users/couet/Downloads/M_Q_analysisSample.root does not exist

As @dastudillo investigated further I did not look more

David_Vannerom · June 1, 2022, 9:25am

It’s weird the mass M and charge Q are not followed by their values. It should be run as:

python makeSelectionEfficiency.py -m 100 -q 1

Also, did you make sure to run the python script from where the files are stored?

I think it would be interesting to see if it works for you as for @dastudillo or if it still overlaps the labels as it does for me.

couet · June 1, 2022, 10:45am

I took the root files from here: Index of /~vannerom/FCP/MuonHack

dastudillo · June 1, 2022, 11:17am

@couet , take those 4 root files and in the code just modify the first line inside def fillEfficiency (line 9), to this:

	fileName = "M"+mass+"_Q"+charge+"_"

Also, you may want to remove the subfolder on the lines (301 and 302) that save the canvas, i.e.:

	c.SaveAs("selectionEfficiency_M"+mass+"_Q"+charge+".png")
	c.SaveAs("selectionEfficiency_M"+mass+"_Q"+charge+".pdf")

and then run as

 python3 makeSelectionEfficiency.py -m 100 -q 1

I get this output too:

Filling file M100_Q1_analysisSample.root
nEntries = 8087
Filling file M100_Q1_muonHack.root
nEntries = 7977
Filling file M100_Q1_elossDown.root
nEntries = 7977
Filling file M100_Q1_elossUp.root
nEntries = 7977
Scaling from the hack after L0 = 1.0166972226447741     -0.0005041034664337372  -0.0003372215904805568
Scaling from the hack after L1 = 1.0187046428657727     -0.0018042889746166452  0.002405718632822149
Scaling from the hack after L1 = 1.0180919636806016     -0.0018053747805538466  0.0024071663740720174
Scaling from the hack after L1 = 1.0180919636806016     -0.0018053747805538466  0.0024071663740720174
Scaling from the hack after L1 = 1.017174993799565      -0.0018080950215144194  0.0024107933620193147
Scaling from the hack after L1 = 1.0144156049208461     -0.002123522469922048   0.0027302431756141194
Scaling from the hack after L1 = 1.005038571051364      -0.0023501922523164556  0.0033574175033092434
Scaling from the hack after L1 = 1.002875572896106      -0.002045345652679255   0.0030680184790188823
Scaling from the hack after L1 = 1.0079357108649998     -0.002399314883133104   0.0034275926901899535
Scaling from the hack after L1 = 1.0082247168540106     -0.003113637901243888   0.004497476968463496
Scaling from the hack after L1 = 1.0145599046132        0.018266817065783956    -0.010231905764404033
Scaling from the hack after L1 = 0.9914592328596252     0.039764070096312186    0.045540328495463744
Scaling from the hack after HLT = 0.992282588975215     0.03797254252297754     0.04302127471719905
Info in <TCanvas::Print>: png file selectionEfficiency_M100_Q1.png has been created
Info in <TCanvas::Print>: pdf file selectionEfficiency_M100_Q1.pdf has been created

couet · June 1, 2022, 11:51am

I get this:

David_Vannerom · June 1, 2022, 12:44pm

Ok then I don’t get how you’re just getting something different. What version of ROOT/python are you using? Also I don’t see the x-axis title on your version of the plot… Maybe in your version the offset is too large.

couet · June 1, 2022, 1:29pm

I am using ROOT master and python3.

David_Vannerom · June 1, 2022, 1:57pm

So I upgraded to ROOT 6.26 (I was on ROOT 6.24) and I now get the same plot as you. So there is something in 6.24 that is such that with a single plot the labels rotate but not anymore with a ratio plot, and this seems to be fixed with 6.26. That’s how I understand this.

Thanks for the cross-checks!