matrix_plot should also accept numpy arrays

[sagetrac-whuss]

Component: graphics

Issue created by migration from https://trac.sagemath.org/ticket/4717

[sagetrac-whuss]

Attachment: matrix_plot.patch.gz

[mwhansen]

Attachment: trac_4717.patch.gz

[mwhansen]

comment:1

Everything looks good except that it's the transpose of what is what wanted. I attached a patch which removes the transpose. I'm not sure if that needs an additional review or not. Wilfried, do you want to take a look?

Apply only trac_4717.patch.

[sagetrac-whuss]

comment:2

With the transpose removed, the following doctest fails,

sage -t  "devel/sage-matrix_plot/sage/plot/matrix_plot.py"  
**********************************************************************
File "/local/data/huss/software/sage-3.2.1/devel/sage-matrix_plot/sage/plot/matrix_plot.py",
line 40:
    sage: list(sorted(m.get_minmax_data().items()))
Expected:
    [('xmax', 4), ('xmin', 0), ('ymax', 3), ('ymin', 0)]
Got:
    [('xmax', 3), ('xmin', 0), ('ymax', 4), ('ymin', 0)]
**********************************************************************

and the rows of matrices are plottet as columns.

Greetings,

Wilfried

[mwhansen]

comment:3

Weird, with the transpose I get that the rows are plotted as columns. For example, with your original patch,

sage: matrix_plot([[1,1,10],[1,1,1],[1,1,1]])

produces http://sage.math.washington.edu/home/mhansen/4717.png .

[sagetrac-whuss]

Attachment: trac_4717.2.patch.gz

final version

[sagetrac-whuss]

comment:4

You are right, the transpose was wrong.
But I also mixed up the xrange and yrange, that was why I got
strange results for non square matrices.

The new patch fixes it. Now matrix_plot behaves the same as
without the patch.

Apply only trac_4717.2.patch

Greetings,

Wilfried.

[sagetrac-mabshoff]

comment:6

Merged trac_4717.2.patch in Sage 3.2.2.alpha1