Rank hotfix

[tjmahr]

The rank-normalization paper says:

Rank normalization proceeds as follows. First, replace each value θ^(nm) by its rank r^(nm) within the pooled draws from all chains. Average rank for ties are used to conserve the number of unique values of discrete quantities.

I was breaking ties by using dplyr::row_number() which works like rank(.., ties.method = "first"). This update changes the method to use averages.

It also fixes a warning from the unit tests.

[codecov-io]

Codecov Report

Merging #199 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #199      +/-   ##
==========================================
+ Coverage   99.35%   99.35%   +<.01%     
==========================================
  Files          30       30              
  Lines        4204     4206       +2     
==========================================
+ Hits         4177     4179       +2     
  Misses         27       27

Impacted Files	Coverage Δ
R/mcmc-traces.R	100% <100%> (ø)	⬆️
R/mcmc-intervals.R	99.24% <100%> (ø)	⬆️

---

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 6dacc42...1499635. Read the comment docs.

[jgabry]

Thanks!