REF: make ipywidgets optional dependency

[martinfleis]

Fixes #130

CI is failing on some numba stuff I did not try to debug.

Btw, we should really try to do #124.

[jGaboardi]

Btw, we should really try to do #124.

Agreed 100%, though I don't want to step on @slumnitz's toes and make changes without her approval.

[jGaboardi]

Seems like the first thing to do would be to switch to main as the default branch and remove master.

[slumnitz]

@martinfleis thank you or taking care of this! Could you add a test as well?

[martinfleis]

I wanted to but we would need to revise CI to test in the environment with and without ipywidgets which doesn't seem to be possible now.

[slumnitz]

@martinfleis ok, most new CI testing happened in https://github.com/pysal/splot/tree/main I am currently not sure anymore why this branch failed. I unfortunately do not have much time this week. I think I could try giving it a go Wednesday during the day, and do testing of this later.

[martinfleis]

I will add the test here and run it locally. Then we can have a go at #124 with @jGaboardi with your permission and ensure that it is properly tested on CI. That way we can merge this and release it before the feature freeze.

[jGaboardi]

In order to complete #124 we will need full admin privileges (for Secrets, etc.).

[martinfleis]

It may actually be overly complicated to do this test withoutpytest (read I don't know how to do it with nose). Can we make a note in #124 or open an issue and add this test later?

[sjsrey]

Just a note that yesterday we removed support for 3.6 and added 3.9 support for libpysal. So if 3.6 is failing in splot, this is likely one reason.

[sjsrey]

In order to complete #124 we will need full admin privileges (for Secrets, etc.).

@jGaboardi and @martinfleis now have admin rights.

[martinfleis]

CI is failing since March 1 (https://travis-ci.org/github/pysal/splot/builds/760938486).

[slumnitz]

@martinfleis & @jGaboardi we used to start fixing all bugs on the main branch, not the master branch anymore and got that one ready including a new testing procedure. The main branch currently seems to pass, however I am unsure if this is correct. Maybe @jGaboardi you could confirm? if so, we should make main the new default and point all new pr's towards main instead of master.

[martinfleis]

I didn't know we have both master and main (confusing). Should I target this PR to main then?

[slumnitz]

@martinfleis we were in the middle of migrating to main and also of during this process changing the testing procedure. However as far as I remember the testing did not work and we could not figure out why. Help is much appreciated. Maybe target it towards main and we see if testing there runs through.

[martinfleis]

let me clean the PR, changing base caused a mess

[martinfleis]

cool, it seems that the new CI is up and running! I'll add the test then and remove ipywidgets from one of the env files to check it.

[martinfleis]

Though it doesn't seem to work...

[slumnitz]

I think it fails due to libpysal not working with 3.6. Can you add the correct python versions within this PR? Just dropping 3.6 adding 3.9.

[slumnitz]

ok now, for some reason it is testing _viz_bokeh.py which is not user exposed and by now likely deprecated (a remain of GSOC 2018 remaining for completeness and potentially for future PR's) is there a way to define not to test this functionality?

[martinfleis]

I added skips for bokeh tests.

[martinfleis]

And test of ipywidgets import error.

[codecov-commenter]

Codecov Report

Merging #132 (686d138) into main (077a5b8) will decrease coverage by 12.20%.
The diff coverage is 75.00%.

@@             Coverage Diff             @@
##             main     #132       +/-   ##
===========================================
- Coverage   98.15%   85.94%   -12.21%     
===========================================
  Files          18       18               
  Lines        1195     1217       +22     
===========================================
- Hits         1173     1046      -127     
- Misses         22      171      +149     

Impacted Files	Coverage Δ
splot/tests/test_viz_bokeh.py	33.33% <31.57%> (-66.67%)	⬇️
splot/_viz_giddy_mpl.py	94.70% <100.00%> (+0.07%)	⬆️
splot/tests/test_viz_giddy_mpl.py	100.00% <100.00%> (ø)
splot/_viz_bokeh.py	16.50% <0.00%> (-81.56%)	⬇️
splot/_viz_utils.py	56.56% <0.00%> (-37.38%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 077a5b8...686d138. Read the comment docs.

[jGaboardi]

I think this is good to merge