[yt-users] Overlaying matplotlib plots on yt plots

[Nathan Goldbaum]

On Mon, Jun 1, 2015 at 9:13 PM, DORIS LEE <dorislee at berkeley.edu> wrote:

> Hi all,
>
> I have been trying to plot a circle of a given radius around the boundary
> of a collapsing cloud to track its evolution.
> Usually in matplotlib, I would just draw a circle using a xy meshgrid, but
> if I do this using matplotlib the circle would be on a separate figure and
> not overlaid on the yt plot. So I tried using a PlotWindow to separate the
> data and the yt plot  but I get this error:
>
> In [92]:yt.visualization.plot_window.PlotWindow(proj,(-1e5,1e5,-1e5,1e5))
>
> AttributeError: 'ProjectionPlot' object has no attribute '_key_fields'
>
>
PlotWindow is a base class for the other plot objects, and it should never
be instantiated directly.  We could probably add some logic so the error
here is less obscure just in case a future user gets confused and tries to
do this.


> In general, is there to get my RAMSES data loaded into yt as a numpy array
> with actual numbers in them then feed it into matplotlib to do this task?
> or alternatively, is there a simpler way to do this directly on yt using
> annotate_contour or some other yt function?
>
> Thank you.
>
> Regards,
> Doris
> ----
> Code for generating the attached plot, the green boundary line is in
> theory what I am trying to draw.
>
> ds= yt.load("../../bin/output_00002/info_00002.txt")
> proj = yt.ProjectionPlot(ds, "y", "density",window_size=10)
> proj.set_cmap("density","rainbow")
> proj.annotate_velocity()
> proj.show()
>

Have you seen the sphere callback?

http://yt-project.org/docs/dev/visualizing/callbacks.html#overplot-a-circle-on-a-plot

You can also access the matplotlib figure and axes object that the
ProjectionPlot wraps:

http://yt-project.org/docs/dev/cookbook/simple_plots.html#accessing-and-modifying-plots-directly

Additionally, if you want the raw data to create the plot yourself using
matplotlib, plot objects have an "frb" attribute (short for
FixedResolutionBuffer) you can use to get the raw image you can pass to
matplotlib's imshow function.  In the above script, you should be able to
do something like:

    dens = proj.frb['density']

    print dens.shape
    print dens.units

    # strip units to pass to matplotlib
    dens = np.array(dens)

    plt.imshow(dens)

You can also create a FixedReslutionBuffer without creating a plot object
first:

http://yt-project.org/docs/dev/quickstart/data_objects_and_time_series.html?highlight=frb

(this is actually how the PlotWindow class is implemented internally).

Hope that's helpful,

Nathan


> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.spacepope.org/pipermail/yt-users-spacepope.org/attachments/20150601/712c5b16/attachment.htm>