That's great news -- glad we got to the bottom of it, though I'm not
sure how your system may have become wedged like that in the first
place. I should have thought of this earlier, but if it happens again,
can you send me your fontList.cache file so I can inspect it? There may
be a bug in th
Hmm... I'm a bit stumped. Can you print out the values of these from
your script, i.e. put the following at the top:
from matplotlib import rcParams
print rcParams['mathtext.fontset']
print rcParams['mathtext.default']
Can you try deleting your fontList.cache file?
Mike
william ratclif
>From my mpl-data directory, here's what's in the mathtext section of my
matplotlibrc file:
# The following settings allow you to select the fonts in math mode.
# They map from a TeX font name to a fontconfig font pattern.
# These settings are only used if mathtext.fontset is 'custom'.
# Note that
The puzzling thing is this:
u'C:\\WINDOWS\\Fonts\\HTOWERTI.TTF'
It's using a custom font in mathtext. Are you setting the rcParams
mathtext.fontset or mathtext.default? That may the culprit, and if not,
it's a bug that it's trying to use that font.
Mike
william ratcliff wrote:
> On the plus
On the plus side, there is no longer an error when I apply the patch. On
the downside, it generates a rather strange symbol instead of a
perpendicular symbolLet me try to quickly upgrade to 0.99.1. I did that
and I seem to get the same error...
On Mon, Apr 19, 2010 at 10:49 AM, Michael Droet
Does forcibly casting the path to a string resolve the problem? i.e.
applying this patch:
Index: mathtext.py
===
--- mathtext.py (revision 8216)
+++ mathtext.py (working copy)
@@ -597,7 +597,7 @@
cached_font = self._fonts.
Assuming that the matplotlibrc file being read is in mpl-data, here is the
relevant section:
### FONT
#
# font properties used by text.Text. See
# http://matplotlib.sourceforge.net/api/font_manager_api.html for more
# information on font properties. The 6 font properties used for font
# matchin
Mike,
The basename is:
u'C:\\WINDOWS\\Fonts\\HTOWERTI.TTF'
Let me try to find where my matplotlibrc file is located...
Thanks,
William
On Mon, Apr 19, 2010 at 10:22 AM, Michael Droettboom wrote:
> One might see that error if the path to the font being used contains
> non-ascii characters (the
One might see that error if the path to the font being used contains
non-ascii characters (the "basename" variable in the last frame of the
stack in the stacktrace). Is that possible? We may need to implement
the same workaround we use for image files for loading fonts (which is
to open the f
I'm not able to reproduce that here, with either SVN or 0.99.1.2. Do
you have any font-related or mathtext-related settings in your matplotlibrc?
Mike
william ratcliff wrote:
> I think the actual error was:
> TypeError: cannot return std::string from Unicode object
>
> It was the error returned
I think the actual error was:
TypeError: cannot return std::string from Unicode object
It was the error returned when I walked through with a debugger...
On Mon, Apr 19, 2010 at 8:51 AM, Michael Droettboom wrote:
> It looks like the end of the traceback -- where the actual exception is
> named
It looks like the end of the traceback -- where the actual exception is
named -- is missing. Can you repost it in its entirety?
Mike
william ratcliff wrote:
> Hi! I am using matplotlib 0.99.0 under windows xp. I tried the
> following:
> ax.text(.96,.80,r'$P \perp
> Q$',fontsize=18,horizonta
Hi! I am using matplotlib 0.99.0 under windows xp. I tried the following:
ax.text(.96,.80,r'$P \perp
Q$',fontsize=18,horizontalalignment='right',verticalalignment='top',transform=ax.transAxes,color='black')
and get the following error:
(However, in figure labels, symbols using mathtext, such as
13 matches
Mail list logo
