@interact

[William Stein]

Hi,

I decided to try just a single random example of @interact in Jupyter
with the sage kernel, since I guess Jereon (?) built
an implementation of interact on top of ipywidgets, and I was
surprised that it failed pretty badly. If you give this as input,

@interact
def f(A = matrix([[1,1],[-1,1]])):
print(A)
print(type(A))

# the output is:
(1, 1)
<type 'sage.modules.vector_integer_dense.Vector_integer_dense'>

So it thinks that matrix is a vector...

I don't think ipywidgets even has a matrix input control, so I'm
curious what the plan was here? In any case, I guess @interact is
still maybe missing some things compared to what is in sagenb? I
couldn't find anything in trac about this.

-- William

https://trac.sagemath.org/search?q=%40interact+matrix+jupyter


--
William (http://wstein.org)

[Kwankyu]

This gets a grid input in Jupyter. Is this close to what you want?

@interact

def _(A=input_grid(2,2, default=[[1,2],[3,4]])):

    A = matrix(A)

    print(A)

[William]

Thanks!   I see Jereon also created https://trac.sagemath.org/ticket/27735

[kcrisman]

This does (still) work in Sage cell, and probably in sagenb (?).

https://sagecell.sagemath.org/?z=eJxzyMwrSS1KTC7h5UpJTVNI03BUsFXITSwpyqzQiI421DGM1YnWBVGxmppWvFwKQFBQBNSj4aiJzCupLEgFCmkCAJITFfI=&lang=sage&interacts=eJyLjgUAARUAuQ==