Use ApproxFunBase.mean instead of Statistics

[jishnub]

This removes the explicit dependency on Statistics, and these may differ after JuliaApproximation/ApproxFunBase.jl#506

[codecov]

Codecov Report

Patch coverage: 91.26% and project coverage change: +22.65% 🎉

Comparison is base (75383fe) 49.48% compared to head (9a16411) 72.13%.
Report is 8 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff             @@
##           master     #281       +/-   ##
===========================================
+ Coverage   49.48%   72.13%   +22.65%     
===========================================
  Files          28       29        +1     
  Lines        2696     2738       +42     
===========================================
+ Hits         1334     1975      +641     
+ Misses       1362      763      -599     

Files Changed	Coverage Δ
src/ApproxFunOrthogonalPolynomials.jl	87.80% <ø> (+28.62%)	⬆️
src/Spaces/IntervalSpace.jl	85.71% <ø> (+42.85%)	⬆️
src/ultraspherical.jl	59.74% <82.85%> (+9.74%)	⬆️
...c/Spaces/Ultraspherical/UltrasphericalOperators.jl	83.33% <87.50%> (+43.24%)	⬆️
src/Spaces/Chebyshev/Chebyshev.jl	63.12% <92.00%> (+12.42%)	⬆️
ext/ApproxFunOrthogonalPolynomialsStaticExt.jl	95.55% <95.55%> (ø)
src/Spaces/PolynomialSpace.jl	93.41% <100.00%> (+25.34%)	⬆️
src/Spaces/Spaces.jl	100.00% <100.00%> (+31.50%)	⬆️
src/Spaces/Ultraspherical/Ultraspherical.jl	86.00% <100.00%> (+60.13%)	⬆️
src/fastops.jl	100.00% <100.00%> (ø)
... and 1 more

... and 17 files with indirect coverage changes

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