vector plot should have optional "start" argument

[jasongrout]

It would be really nice if this plotted an arrow from (1,2) to (3,4):

sage: v=vector([1,2])
sage: u=vector([2,2])
sage: plot(u,start=v)

or maybe the option should be "base" or "origin"

To fix this, just change the plot method in devel/sage/sage/modules/free_module_element.pyx

---

Apply attachment: trac_9962_vector_start.2.patch and attachment: trac_9962-reviewer.patch

CC: @sagetrac-ryan @kcrisman

Component: graphics

Author: Ryan Grout, Karl-Dieter Crisman

Reviewer: Aly Deines, Karl-Dieter Crisman

Merged: sage-4.7.1.alpha2

Issue created by migration from https://trac.sagemath.org/ticket/9962

[jasongrout]

comment:1

I think "start=list/tuple/vector" is the best convention for this option.

[sagetrac-ryan]

tentative patch

[sagetrac-ryan]

comment:2

Attachment: trac_9962_start_vector.patch.gz

the patch gives the described output, but someone should double check my code for correctness.

[adeines]

comment:3

Replying to @sagetrac-ryan:

the patch gives the described output, but someone should double check my code for correctness.

It looks good to me.

[jdemeyer]

Author: Ryan Grout

[jdemeyer]

Reviewer: Aly Deines

[sagetrac-ryan]

updated patch

[sagetrac-ryan]

comment:5

Attachment: trac_9962_vector_start.patch.gz

[sagetrac-ryan]

updated patch + doctests

[sagetrac-ryan]

comment:6

Attachment: trac_9962_vector_start.2.patch.gz

updated patch:

Included error handling for some cases where the two coordinates are not of the same dimension. Also, added doctests.

Much cleaner patch.

I appreciate the suggestions for making this patch better.

[adeines]

comment:8

Replying to @kcrisman:

[jdemeyer]

comment:9

Please make it clear which patches have to be applied.

[kcrisman]

comment:11

You may want to check whether this really does the same thing as v.plot() - see #10638.

[jdemeyer]

comment:12

Replying to @kcrisman:

You may want to check whether this really does the same thing as v.plot() - see #10638.

I am interpreting this comment as a needs_work (if it's not, then please set positive_review).

[kcrisman]

comment:13

No, it's neither positive review nor needs work. I haven't had the time to review this, but any reviewer should be sure to check this out. That's all I am saying.

[kcrisman]

Changed reviewer from Aly Deines to Aly Deines, Karl-Dieter Crisman

[kcrisman]

comment:14

This patch causes an amusing error, which does not occur in the vanilla Sage:

sage: plot(vector([1]))
ERROR: An unexpected error occurred while tokenizing input
The following traceback may be corrupted or invalid
The error message is: ('EOF in multi-line statement', (2, 0))

---------------------------------------------------------------------------
ArithmeticError                           Traceback (most recent call last)
<snip>
ArithmeticError: Cross product only defined for vectors of length three or seven, not (3 and 1)

I think I can fix this, so patch hopefully coming up.

[kcrisman]

comment:15

I think I have it fixed. However, the bizarre error messages remain for one-dimensional arrows, so I have created followup ticket #10925 for that.

[kcrisman]

Attachment: trac_9962-reviewer.patch.gz

Apply after vector_start.2 patch

[kcrisman]

Changed author from Ryan Grout to Ryan Grout, Karl-Dieter Crisman

[kcrisman]

comment:16

Okay, positive review on this nice addition from the original patch.

The reviewer patch still needs review; it fixes the problem by extending the one-dimensional start as well, and adds/spruces up some documentation. Reviewer should check things work, doctests, and that documentation looks ok.

[kcrisman]

comment:17

Ryan, you can totally review the reviewer patch. Just check that it still does what you wanted, that the doctests are correct and pass, and that ./sage -docbuild reference html looks nice in the documentation for free modules.

[sagetrac-ryan]

comment:18

Replying to @kcrisman:

Ryan, you can totally review the reviewer patch. Just check that it still does what you wanted, that the doctests are correct and pass, and that ./sage -docbuild reference html looks nice in the documentation for free modules.

I'll look at it this evening.

[sagetrac-ryan]

comment:19

everything looks good here. I'm going to go ahead an change this to positive review for the reviewer patch.

[jdemeyer]

Merged: sage-4.7.1.alpha2