Option names
Ok, I see your point now. It's just the help header that is wrong. Use the correct parameter names (as in the parameter structure line 469) to overcome this. I'll update a corrected version when time allows.
Dot scaling:
When dealing with ':' lines, the size parameters are 'linemode', 'linewidth', 'linemin' and 'linemax'. I've done a test and it seems to work correctly ('-' and ':' lines are scaled identically). Send me an example if you still think there is a problem here.
Yep, I am still around, but I don't have much time at the moment. Concerning the last comments:
tiff in black and white:
- I've never tested it, it must be in the gs conversion from eps to tiff. You should check what eps2xxx does. It can be avoided, if really necessary, by avoiding the eps altogether (see the help header of fig2public).
latex interpreter:
- yep, I tried to go around the bug in Matlab as best as I could. It works for simple expressions. For the rest, I guess you have to play around with your figures manually.
simple example:
- there are examples in the help header of fig2public. Please read them before commenting...
Zi-An Li:
Thanks for the comment! I leave the corresponding rating to my general incapacity to understand things.
Option names:
I don't see why.
'dot' scaling:
I am not sure to understand your point. Which kind of lines are you speaking of? A pure line (i.e. ':') or a mixture of marker and line (e.g. '.-'). According to your comment, I have the feeling that you're speaking of the 2nd situation. In that case, it's not a bug. It's just that lines and markers are two different things. You therefore have to set the options for both types individually.
Thanks for the feedback. Standard options go back to the very first version and should probably be revisited. When I have time, I'll try to find an elegant solution. The best choice of options depends on the kind of figures (multiple axes, log-scales, etc), the figure usage (reports, articles, presentations, etc) and personal tastes (that's a very long list here...). What I usually do is that I put all the most common options in a structure and I call fig2public on different subsets of figures, giving the common option structure as input and redefining the options particular to a set of figures within the calls themselves. By the way, using the 'tight' option is meaningful only if you also use 'eps_loose', otherwise the exported figures should already be quite tight (but the exact figure dimensions not controlled).
If you find a bug, please let me know. It's always worth reading the help headers though.
Thanks for your interest. Can't remove your ratings myself, sorry. For your question:
All the formatting is done via the function formatfig.m. All its options are reachable directly in the call to fig2public. So, you get the option that you need with:
>>help formatfig
Then you use the 'param_name',param_value input (ie in addition to your other options) to set its value, for example:
>>fig2public( ..., 'linemode','fixed','linewidth',2, ... )
Thanks Jon. yes Lube is my real name. When I get around to processing whichever line plots were causing issues I'll get back to you. Oh wait, I just remembered! If you try plotting a line through some data points so that the point markers are shown as well (i.e. '.-'), then the way the dots are scaled means they are hidden behind the lines. Hence the need to scale the markers manually.
There's also another issue that I was thinking of where the dots from a dotted line (i.e. ':') don't come out square but as rectangles, but i think this is an issue with the eps conversion (i.e. not your fault!).
I also just discovered an issue with the 'painters' eps renderer for surface plots, which is that black lines overlaying the plots seem to get randomly converted to other colours. Probably best to use the openGL renderer for surface plots (you might want to update the header options in fig2public to include opengl.). For the benefit of others I found this article helpful in understanding the difference between the renderers: http://www.mathworks.co.uk/support/tech-notes/1200/1201.html#Section_2
Option names
Ok, I see your point now. It's just the help header that is wrong. Use the correct parameter names (as in the parameter structure line 469) to overcome this. I'll update a corrected version when time allows.
Dot scaling:
When dealing with ':' lines, the size parameters are 'linemode', 'linewidth', 'linemin' and 'linemax'. I've done a test and it seems to work correctly ('-' and ':' lines are scaled identically). Send me an example if you still think there is a problem here.
Hi Jonathan,
Thanks for getting back on this. Appreciate that you are busy.
Option names: compare lines 253 with 469 and 470 of formatfig.m, and the same for lines and markers. As it is the string comparison obviously fails. Trivial yet important!
Dot scaling: I was referring to the former i.e. ':'. The dots are incorrectly scaled. I think this is the same problem as discussed here: https://sites.google.com/site/oliverwoodford/software/export_fig.
When you get a minute please have a go at working out these bugs as this really is the best software package out there for preparing figs for papers!
Thanks for all your work.
Cheers.
Yep, I am still around, but I don't have much time at the moment. Concerning the last comments:
tiff in black and white:
- I've never tested it, it must be in the gs conversion from eps to tiff. You should check what eps2xxx does. It can be avoided, if really necessary, by avoiding the eps altogether (see the help header of fig2public).
latex interpreter:
- yep, I tried to go around the bug in Matlab as best as I could. It works for simple expressions. For the rest, I guess you have to play around with your figures manually.
simple example:
- there are examples in the help header of fig2public. Please read them before commenting...
Zi-An Li:
Thanks for the comment! I leave the corresponding rating to my general incapacity to understand things.
Option names:
I don't see why.
'dot' scaling:
I am not sure to understand your point. Which kind of lines are you speaking of? A pure line (i.e. ':') or a mixture of marker and line (e.g. '.-'). According to your comment, I have the feeling that you're speaking of the 2nd situation. In that case, it's not a bug. It's just that lines and markers are two different things. You therefore have to set the options for both types individually.
Another Bug:
the options for fontsize and markersize in formatfig should be specified as:
'fontmin', 'fontmax' etc. as opposed to 'fontsizemin' and 'fontsizemax'.
Also default behaviour for a 'dotted line' plot causes the dots to be scaled smaller than the lines. Can't figure out why, use 'markermode','scaled','markersize',2 as a work-around.
It seems the author is no longer supporting this, which is a shame as this is an excellent package...
If anyone knows how to get in touch with him, then please do so!
Cheers.
Thanks Jon. yes Lube is my real name. When I get around to processing whichever line plots were causing issues I'll get back to you. Oh wait, I just remembered! If you try plotting a line through some data points so that the point markers are shown as well (i.e. '.-'), then the way the dots are scaled means they are hidden behind the lines. Hence the need to scale the markers manually.
There's also another issue that I was thinking of where the dots from a dotted line (i.e. ':') don't come out square but as rectangles, but i think this is an issue with the eps conversion (i.e. not your fault!).
I also just discovered an issue with the 'painters' eps renderer for surface plots, which is that black lines overlaying the plots seem to get randomly converted to other colours. Probably best to use the openGL renderer for surface plots (you might want to update the header options in fig2public to include opengl.). For the benefit of others I found this article helpful in understanding the difference between the renderers: http://www.mathworks.co.uk/support/tech-notes/1200/1201.html#Section_2
Cheers.
Hi Lube (man, what kind of a name is that? ;-))
Option names
Ok, I see your point now. It's just the help header that is wrong. Use the correct parameter names (as in the parameter structure line 469) to overcome this. I'll update a corrected version when time allows.
Dot scaling:
When dealing with ':' lines, the size parameters are 'linemode', 'linewidth', 'linemin' and 'linemax'. I've done a test and it seems to work correctly ('-' and ':' lines are scaled identically). Send me an example if you still think there is a problem here.
Thanks for your comments!
Hi Jonathan,
Thanks for getting back on this. Appreciate that you are busy.
Option names: compare lines 253 with 469 and 470 of formatfig.m, and the same for lines and markers. As it is the string comparison obviously fails. Trivial yet important!
Dot scaling: I was referring to the former i.e. ':'. The dots are incorrectly scaled. I think this is the same problem as discussed here: https://sites.google.com/site/oliverwoodford/software/export_fig.
When you get a minute please have a go at working out these bugs as this really is the best software package out there for preparing figs for papers!
Thanks for all your work.
Cheers.
Yep, I am still around, but I don't have much time at the moment. Concerning the last comments:
tiff in black and white:
- I've never tested it, it must be in the gs conversion from eps to tiff. You should check what eps2xxx does. It can be avoided, if really necessary, by avoiding the eps altogether (see the help header of fig2public).
latex interpreter:
- yep, I tried to go around the bug in Matlab as best as I could. It works for simple expressions. For the rest, I guess you have to play around with your figures manually.
simple example:
- there are examples in the help header of fig2public. Please read them before commenting...
Zi-An Li:
Thanks for the comment! I leave the corresponding rating to my general incapacity to understand things.
Option names:
I don't see why.
'dot' scaling:
I am not sure to understand your point. Which kind of lines are you speaking of? A pure line (i.e. ':') or a mixture of marker and line (e.g. '.-'). According to your comment, I have the feeling that you're speaking of the 2nd situation. In that case, it's not a bug. It's just that lines and markers are two different things. You therefore have to set the options for both types individually.
Cheers
Another Bug:
the options for fontsize and markersize in formatfig should be specified as:
'fontmin', 'fontmax' etc. as opposed to 'fontsizemin' and 'fontsizemax'.
Also default behaviour for a 'dotted line' plot causes the dots to be scaled smaller than the lines. Can't figure out why, use 'markermode','scaled','markersize',2 as a work-around.
It seems the author is no longer supporting this, which is a shame as this is an excellent package...
If anyone knows how to get in touch with him, then please do so!
Cheers.
Comment only