Is this a bug?

Yeah -- certain cases probably have better type safety than others. Long term, we probably want to do more type checking closer to the entry points (rather than deeper where they actually cause exceptions but are at that point long-removed from the user error). That kind of checking doesn't seem to be the norm in languages like Python... but matplotlib is an exception in that it is more tuned to interactive use than many libraries.

Mauro Cavalcanti wrote:

Dear Michael,

Just noticed another quite interesting thing: if I (mistakenly) pass
the "linewidth" parameter as a string (as I was doing previously with
the "markersize" one), no backend error occurs!

2008/12/15 Michael Droettboom <mdroe@...86...>:
  

Can you share the code you used to set the properties of points and lines.
It looks like somehow a string is being set where a floating-point value is
expected (either for dpi or markersize).

Mike

--
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

I feel highly honoured and still happier by having someone from STSI
helping me with my biodiversity mapping project. I appreciate a lot
the STSI Python for interactive data analysis tutorial written by
Perry Greenfield and Robert Jedrzejewki, I have found it very useful!)

Thanks. I'll let them know. I feel highly honored to work with those folks, too... :wink:

Mike

···

--
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA