permutations in feature_importance

[tkonopka]

Updates to function feature_importance following discussion in #29

• argument B, determines number of permutations carried out on each variable. Output object gives average dropout_loss over these permutations
• argument keep_raw_permutations, determines whether or not output object should contain details of all the permutations. Details are provided in attribute raw_permutations.

Comments

• some minor edits to inner code (match.arg)
• testthat chunks for new arguments based on titanic data
• some testthat chunks for miscellaneous items in feature_importance
• output is accepted in plot
• code passes devtools::test()

[codecov-io]

Codecov Report

Merging #31 into master will increase coverage by 0.86%.
The diff coverage is 100%.

@@           Coverage Diff            @@
##           master    #31      +/-   ##
========================================
+ Coverage   84.13%    85%   +0.86%     
========================================
  Files          29     29              
  Lines        1002   1020      +18     
========================================
+ Hits          843    867      +24     
+ Misses        159    153       -6

Impacted Files	Coverage Δ
R/feature_importance.R	100% <100%> (+8.69%)	⬆️

---

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 6717011...ee84291. Read the comment docs.