On Thu, Mar 11, 2010 at 2:41 PM, Craig the Demolishor wrote:
> Hi folks,
> Lately I've been working with some data that is too copious to fit in
> memory, so I've had to write a wrapper for pyplot.hist that bins the data in
> chunks and then draws it like so:
>
> pyplot.hist(x_edges, bins=100, weights=bin_contents,
> histtype='stepfilled', facecolor='g')
>
> However, when I try to set the histogram's y-scale to logarithmic the
> colors get all messed up (see attached). Any ideas? This is w/ matplotlib
> 0.99.0.
>
>
> Thanks in advance.
>
> --cb
>
> # Example script
> import numpy as
> np
> import matplotlib.pyplot as
> plt
>
>
> # generate some data on
> log-scale
> x =
> 10**np.random.uniform(size=1)
>
> # bin the data myself
> bins, xed = np.histogram(x, bins=100, range=(0,
> 10))
>
> # plot the data as a weighted histo
>
>
> plt.hist(np.linspace(0, 10, 100), bins=100, weights=bins,
> histtype='stepfilled',
> facecolor='g')
>
> plt.yscale('log')
>
> plt.savefig("test.png")
>
> plt.clf()
>
>
>
Use log=True in the plt.hist instead of explicitly setting the in the
yscale()
>
> --
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> ___
> Matplotlib-users mailing list
> Matplotlib-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
--
Gökhan
--
Download Intel® Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev___
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users
