On 2012/08/13 3:58 PM, Benjamin Root wrote:
>
>
> On Thu, Aug 2, 2012 at 7:21 PM, Chad Parker <parker.char...@gmail.com
> <mailto:parker.char...@gmail.com>> wrote:
>
> Hello everyone-
>
> I'm a new Python/Matplotlib user, but I have quite a bit of plotting
> experience with octave/matlab and gnuplot. So, I apologize in
> advance if my python style is terrible and if I give you all the
> wrong information!
>
> I'm having a problem saving figures that contain images as eps files
> when my x-axis is a date/time axis. The trouble is that in the
> resulting eps file, the image data is shifted relative to the axis.
> I've attached two images as an example. In the .png file the
> alignment is correct and in the eps file it's not. The png is also
> nearly identical to what I get when I plot to the screen. The script
> I wrote to generate these plots is at the end of this message.
>
> It's worth noting that the two output images are identical if the
> x-axis is simply numerical and not a time series.
>
> I'm using python 2.7.3 and Matplotlib 1.1.0
>
> Some Google searches dug up this old thread, however, the shift that
> I'm experiencing seems to be much greater, so I don't know if it's
> related:
> http://old.nabble.com/Saving-as-eps-file-shifts-image--td29232680.html
>
> Can anyone help me solve this?
> Thanks in advance,
> --Chad
>
>
> Here is a minimal script that I used to generate these plots:
>
> import matplotlib.pyplot as plt
> import matplotlib.dates as dts
> from matplotlib.colors import LogNorm
>
> imgData = [[1.0/x + 1.0/y for x in range(1,100)] for y in range(1,100)]
>
> tMin=dts.epoch2num(1343947266)
> tMax=dts.epoch2num(1343947266+100)
>
> axImg=plt.subplot(111)
> axImg.imshow(imgData, norm=LogNorm(),extent=(tMin,tMax,1,100),\
> interpolation='none', origin="upper")
> axImg.xaxis_date()
> axImg.set_aspect('auto')
>
> plt.savefig("imageshift.png")
> plt.savefig("imageshift.eps")
>
>
> Confirmed. This is a fairly nasty-looking result. And I am not exactly
> sure what is happening here. Note that it does not matter if the png or
> the eps images were saved first. It looks like an auto-scaling bug, but
> given that it doesn't impact the png image, I am not so sure.
This appears to be some sort of truncation error bug; it has nothing to
do with dates. Here is a modified illustration:
import matplotlib.pyplot as plt
from matplotlib.colors import LogNorm
imgData = [[1.0/x + 1.0/y for x in range(1,10)] for y in range(1,10)]
tMin = 1000000066 # Shifted by half
tMin = 100000066 # Blank!
tMin = 10000066 # OK
tMax = tMin+10
# The following combination is also shifted by half
tMin = 10000066.1 # adding the 0.1 doesn't make a difference
tMax = tMin+1
print tMin, tMax
axImg=plt.subplot(111)
axImg.imshow(imgData, norm=LogNorm(), extent=(tMin,tMax,1,10),
interpolation='none', origin="upper")
axImg.set_aspect('auto')
plt.savefig("imageshift.png")
plt.savefig("imageshift.eps")
plt.savefig("imageshift.pdf")
I thought it might come from the use of str() instead of repr() when
generating the concat matrix in the ps backend, but that's not it. I'm
suspecting it may be inherent in the ps language, when one uses a matrix
to make an enormous translation in one direction, and then uses
translate to sling everything back in the other direction. Either that,
or some sort of "snapping" is going on.
Eric
>
> Could you file a bug report on this at the github tracker? I would like
> to see this fixed before the 1.2.0 release.
>
> Thanks!
> Ben Root
>
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
