Dear Experts,
I’m facing some strange issue, where my TGraphErrors drawn with option “L4” draws some very strange error bands. They look to me inverse or shifted. Unfortunately I couldn’t figure out why. My first guess was a negative error causing this behaviour but using an additional “abs()” to set the error didn’t solve the issue.
Here’s a screenshot of the behaviour:
Feel free to ask for additional info if needed!
Thanks
Harald
ROOT Version: 6.12.06
Platform: slc6
Compiler: gcc62
Do you have a macro reproducing the issue ? Are the x values in increasing order ?
I stripped down my script via commenting code to run as a macro and added a section where I set the options I’d normally pass via commandline. You just have to call “root graph.cxx” and it produced a “test_fit_reweightMC_01_wNone.pdf” in the folder you run it.
graph.cpp (12.6 KB)
I can remove the unnecessary code (commented out) if you wish.
Thanks
Harald
Edit: I forgot; yes the x-values are in increasing order.
Ok using “L3” to draw the TGraphErrors the plot looks fine. So something sketchy must be happening when using “L4” which shall draw a smoothed filled area.
I see. You are facing the issue described here. The smoothing technique used when option “E4” is set may produce some artefacts. Yes, in your case, you should go to 3.
Ok, must have overread ‘artefacts’ there. So I’ll go with ‘E3’ instead.
Thanks
Harald
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.
