Introduce class attr BaseViz.enforce_numerical_metrics

[mistercrunch]

@michellethomas

[codecov-io]

Codecov Report

Merging #5176 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #5176      +/-   ##
==========================================
+ Coverage   77.46%   77.47%   +<.01%     
==========================================
  Files          44       44              
  Lines        8729     8732       +3     
==========================================
+ Hits         6762     6765       +3     
  Misses       1967     1967

Impacted Files	Coverage Δ
superset/viz.py	81.33% <100%> (+0.03%)	⬆️

---

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 a3477ab...c3501b2. Read the comment docs.

[michellethomas]

lgtm This seems like a good solution to me. Ideally we'd be getting the correct types in pandas and wouldn't need to coerce to num at all, but when I looked around I didn't find a great solution to this problem.

When were you seeing the incorrect object types (what are the steps to reproduce)? Just for future reference it would be nice to better understand the problem.

[michellethomas]

@mistercrunch this lgtm, merge?

[mistercrunch]

Related unmerged PR here:
#4724