I am going to start using GNUPlot for my future papers, because it is perfectly integrated with LaTeX and Tcl. Recently I used Matplotlib (Python), R, Plotchart (Tcl), and D3 (JS). I don't like to proceed with them because PDF, SVG, or PS are required as intermediate files and it's not good for the version control. I'd like to ask people with GNUPlot/LaTeX hands-on experience what are the disadvantages/bottle-necks of GNUPlot??? Please share your thoughts on it
I don't really see why intermediate files like PDF, SVG, or PS are a problem. Usually the publisher wants to have all of your images in one of these formats anyway, even when you submit your manuscript as LaTeX/pdf. And for version control you can use Git for example.
That said, for me the biggest bottleneck of GNUPlot is finding the right "terminal" that has all the functionality/appearance that I want. Furthermore, in my opinion the documentation of matplotlib is much easier to understand than GNUPlot's documentation.
I am a bit of a purist and therefore value high quality documentation and figures. That is why I started using LaTeX and gnuplot in 2013. I am still using LaTeX today, but have not used gnuplot for several years now. I mostly generate my charts with Matplotlib, and if I am REALLY pressed for time with Excel (I hate to admit).
I use Matplotlib mostly because I can code my calculations and charting all in one file. It's just faster to get it done than writing a separate awkwardly syntaxed gnuplot script.
Having said that, if I had to prepare figures for a journal publication written in latex, I would return to gnuplot in an instant. The quality is simply unrivalled.
All the best
Johan
Since it is no problem to submit LaTeX files to a journal with embedded graphics or plots, the GNUPlot package seems to be a nice addition. The visualization of different data files is quite sufficient and it is possible to run data fits within the GNUPlot environment. From a scientific point of view I would say it is fine.
Out of my experience I can say that I prefer GNUPlot for data fitting and simple visualization because it is easy to write code for that. But I usually do not use the plots within the LaTeX files I write publications with. GNUPlot has some limitations when it comes to complex data analysis or formatting the resulting data for which additional LaTeX packages might be needed or for which Python with the corresponding packages might be better.
I would say it depends on the complexity of what you are going to analyze which programs or which combinations are best.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python