Jae-Joon Lee wrote:
> On Wed, Jan 20, 2010 at 2:04 PM, Eric Firing <efir...@hawaii.edu> wrote:
>> The reason for this fudge in contour is that contourf fills
>> lower < z <= upper
>> for each consecutive pair of contour levels.
>> When the minimum value of z coincides with the lowest level, then regions
>> with that minimum are left blank; so the lowest level is adjusted downward
>> slightly, making the lowest contour interval include the minimum value.
>>
> 
> I understand levels can be adjusted for a better contouring, but I'm
> not sure whether this change needs to be visible to users.
> 
> The autoleveler, initially creates following levels
> 
> [0, 1.5, 3, 4.5, 6, 7.5, 9.]
> 
> But due to the reason you described above, they become
> 
> array([ -9.00000000e-06,   1.50000000e+00,   3.00000000e+00,
>          4.50000000e+00,   6.00000000e+00,   7.50000000e+00,
>          9.00000900e+00])
> 
> And the colorbar uses the adjusted levels for labeling. But I think it
> may make more sense to use the initial levels (at least for the
> colorbar ticks).

JJ,

I changed my mind and decided you are correct in thinking the change 
should be made in contour.py.  I now make the bottom boundary adjustment 
at the last possible time, and in such a way that it does not change the 
  levels array at all.  Therefore the colorbar never sees it, and the 
result of explicitly supplying a set of levels is identical to the case 
where those levels result from autoscaling.

Eric
> 
> Regards,
> 
> -JJ


------------------------------------------------------------------------------
Throughout its 18-year history, RSA Conference consistently attracts the
world's best and brightest in the field, creating opportunities for Conference
attendees to learn about information security's most important issues through
interactions with peers, luminaries and emerging and established companies.
http://p.sf.net/sfu/rsaconf-dev2dev
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Reply via email to