Calling the extrema function on a SparseArray falls back to iterating over all elements on the Array (including zeros).
This makes it extremely slow for large sparse matrices. It would be better to compute the extrema of the nonzeros and then account for the zeros (if there are zeros).

Kind regards,
Tom

The last few weeks I have to switched to JuMP in order to perform my (optimization) analysis. Having a look at the documentation related to "Query Solutions" I found that you can extract very useful information regarding the optimizations results.

Thus, I was thinking that it would be a nice all of this information to be presented in a table. I understand that not everything is useful for every

Hi there,

I use Jupyter (specifically Jupyter Lab) to make project reports every 3 months or so. As a result, the reports get quite large. I structure them in detail with markdown headers and sections. Jupyter provides an excellent extension, called simply "table of contents" that does two things:

1. Enables a table of contents sidebar that is clickable and brings you to the markdown hea

Only relevant to pyplot with most recent matplotlib

plot(1:10)
sys:1: MatplotlibDeprecationWarning: Passing the fontdict parameter of _set_ticklabels() positionally is deprecated since Matplotlib 3.3; the parameter will become keyword-only two minor releases later.
sys:1: UserWarning: FixedFormatter should only be used together with FixedLocator
sys:1: MatplotlibDeprecationWarning: Case-i

I found that function mod2pi is not implemented yet, but mod works. Is there any list of implemented functions? Minimal working example is:

using Zygote

# This is working
gradient(x -> mod(x, 2pi), 1.)

# This is not
gradient(x -> mod2pi(x), 1.)

So that other people (like me) can help development of the interactive API. At the moment it is something alien that I Can just use through example. This is not the proper way forward...

(finally found Mouse in interaction/iodevices.jl...)

Hi there. It's my first day coding & I really appreciate this resource and I find it more or less easy to follow. However, I would recommend either providing a link to a glossary or simply having a glossary available for absolute beginners to explain terminology such as those referring to the various data types & examples given. Just to make the language a bit more accessible.

I'm enjoying

It would be handy for all edge operations to have both f(g, e) and f(g, src(e), dst(e)) defined by default on abstract graphs