# global width variable scaled by factor 0.95092 #672

Closed
opened this issue May 26, 2015 · 1 comment

Projects
None yet
2 participants

### Jay-At-Play commented May 26, 2015

Member

### egeerardyn commented May 26, 2015

 The 0.95... comes from the way we currently handle subplots. This is already reported in #622, where I wrote a more extensive explanation of what happens and how you can temporarily patch your copy of matlab2tikz to mitigate this issue. We haven't fixed it yet, since we have to account for how different people use this feature: some specify no widths, some only height, some use absolute dimensions (5cm) and some use semantic ones (\figwidth). All are slightly different, but we try to keep those in the back of our heads. Unfortunately, we didn't do that for that semantic dimensions the last time we changed things. However, in your case, I also noticed that you are defining \figwidth as a macro (using \newcommand or \def), not as a length. That is not a recommended approach, if you use \newlength and \setlength instead as on that wiki page, you don't need any work-arounds in your LaTeX code: 0.95\figwidth will mean what you expect. Since this is a duplicate of #622, I will be closing this issue report. If you have have further questions regarding my response, please feel free to respond here. If you want to discuss that feature further, please do so in #622.