Use approximate sparsity detection by default

[avik-pal]

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the ScioML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

If Symbolics.jl is not loaded, it uses ForwardDiff for approximate sparsity detection. I am displaying a warning because this can lead to failures due to incorrect sparsity pattern. But overall this is very promising at least for the brusselator example. Approximate Detection + Solving takes the similar time as Symbolics Solving (symbolic detection being extremely slow)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (d7ef4af) 80.48% compared to head (893d084) 89.37%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #318      +/-   ##
==========================================
+ Coverage   80.48%   89.37%   +8.89%     
==========================================
  Files          23       24       +1     
  Lines        1942     1949       +7     
==========================================
+ Hits         1563     1742     +179     
+ Misses        379      207     -172     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.