Sorry, had completely forgotten this. Could you post the original PDF file ? I think this may be a but in the PDF rendering rather than in ctioga2 (at least, I can't reproduce it with data looking reasonably like yours).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hmmm. On second thougts, I don't know. The line is visible in Acrobat reader and in xpdf. The line keeps getting thinner when I zoom in, so it's not a "real" line... How does it print ?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I also don't see the line when opening with evince.
But it is there when using okular or adobe acrobat reader.
Now I printed the pdf from evince and there is printed thin blue line connecting the bars (y=0).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
In the PDF file, I can clearly see that the last operation in the drawing of the histograms should be a "moveto", but is in fact a "lineto". Changing that by hand seems to work (see attached PDF). I'll see what I can do in the source code, though...
Hey, I've looked into that further, and into the PDF spec. I'm unsure what to do. Clearly, my fix is wrong, and give very bad results in other cases. The problem is that ctioga2 is asking to the PDF viewer to fill a complicated shape that contains a 0-width rectangle, which most viewers chose to render as a very thin line (but some do not).
I'm reverting the fix for now, I'll have to come up with another way of doing that, involving fill each bar separately.
Hmm...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear Vincent,
I plotted my data using filled histogram... And there is an unwanted line (very thin, but visible), y=0...
The code is
xrange -1:14
yrange -0.1:0.2
histogram /gap=0.5
fill 0
fill-color blue
plot data.dat@2:3 /color=black /line-style=no
What do you think?
Thank you in advance!
Last edit: Evgenii 2016-02-23
Any ideas regarding this line?
Sorry, had completely forgotten this. Could you post the original PDF file ? I think this may be a but in the PDF rendering rather than in ctioga2 (at least, I can't reproduce it with data looking reasonably like yours).
Thank you for the reply!
Attached is the pdf file.
Indeed, this is an anti-aliasing artifact of the PDF display. Here's what I have using evince.
Hmmm. On second thougts, I don't know. The line is visible in Acrobat reader and in xpdf. The line keeps getting thinner when I zoom in, so it's not a "real" line... How does it print ?
I also don't see the line when opening with evince.
But it is there when using okular or adobe acrobat reader.
Now I printed the pdf from evince and there is printed thin blue line connecting the bars (y=0).
Hello!
Do you have any news regarding this line?
Last edit: Evgenii 2017-01-01
Wow, sorry for being so long.
In the PDF file, I can clearly see that the last operation in the drawing of the histograms should be a "moveto", but is in fact a "lineto". Changing that by hand seems to work (see attached PDF). I'll see what I can do in the source code, though...
This is fixed in the lastet commit, https://sourceforge.net/p/ctioga2/code/ci/28d3c0529d2dfbd3b8cdc5831254378c4561fc19/
Please try out and tell me !
It works! Thank you!
Hey, I've looked into that further, and into the PDF spec. I'm unsure what to do. Clearly, my fix is wrong, and give very bad results in other cases. The problem is that
ctioga2is asking to the PDF viewer to fill a complicated shape that contains a 0-width rectangle, which most viewers chose to render as a very thin line (but some do not).I'm reverting the fix for now, I'll have to come up with another way of doing that, involving fill each bar separately.
Hmm...