Felix Zoergiebel

I have problems with calling ghostscript (and also pdf2ps) from MATLAB R2011b on OS X 10.6.8. system('/opt/local/bin/gs') gives me this error:

dyld: Library not loaded: /opt/local/lib/libfreetype.6.dylib
  Referenced from: /opt/local/lib/libfontconfig.1.dylib
  Reason: Incompatible library version: libfontconfig.1.dylib requires version 14.0.0 or later, but libfreetype.6.dylib provides version 10.0.0
/opt/local/bin/gs: Trace/breakpoint trap

I fixed the problem by deleting MATLABs DYLD_PATH_LIBRARY variable before calling ghostscript in ghostscript.m:

[varargout{1:nargout}] = system(sprintf('export DYLD_LIBRARY_PATH=""; /opt/local/bin/gs %s', cmd));

Same should work for pdf2ps.m.

This problem does not exist on MATLAB 7.4, since there is no libfreetype.6.dylib in the MATLAB installation. Hope that this little report helps somebody else.

Great job! This really helps me plotting translucent patches, which is just not possible with matlab functions. However, there are small problems.
1: latex \cdot is not processed. I added
tex=strrep(tex,'\cdot','& # 8901;'); (without spaces...)
to line 2351 of plot2svg_beta.m
2: I have an hggroup with text objects that are placed outside of the axis (custom labels). These labels are not plotted before I set the parent of each of them to the axis.

I do not really understand what you want to do, but I guess it's either of the following:
If you want to have errorbars on each of the interpolated points, do

ploterr(x1,y1,0.02,0.5).

If you want the interpolated points plotted as a line and only the "raw data" as points with errorbars, you can do that with two consecutive plot commands:

ploterr(phi1,Iphi1,0.02,0.5,'o')
hold on, plot(x1,y1), hold off

Hope that helps,
Felix

very useful

Hi Felix,
Here's a small mistake catching mechanism that can be added:
I'm using ploterr with the toggle for 'logxy' on, and I didn't realize that there were some points that would be negative with error bars, and I tried plotting them, and ploterr "did" without returning any error. I thought the plot looked a little sparse, and I couldn't figure it out until I finally noticed these tiny pixels on my graph. I'm not quite sure what they are because I haven't looked into ploterr, but it seems that feeding points where data-error<=0 (I tried 0 as well), makes ploterr plot these tiny pixels on your graph instead of the real points you wanted. They're very hard to see, and I missed them the first dozen time I look at my graph; so, it may be handy to provide some kind of error message when that happens.