minor fixes and improvements

notes (Peter's comments before v0.6.0)

• Please fix: I don't understand the description of the eset_final_sub
  data set. What are "theta" and "exprs(est_final)"?

  • fixed

• Please fix: The "fit" data set should be a single object; currently
  it loads two objects, "fit_predict" and "fit_train". Maybe define a
  list with two elements. Also, I suggest using a more descriptive
  name.

  • made into two separate data objects

• Please fix: Why do "fit_bspline" and "fit_loess" return a list with
  only one element? You don't need the list.

  • added pve calculations. so now a list of two vectors.

• Please fix: some undocumented arguments that should be added to the
  roxygen2 docs:

  • Undocumented arguments in cycle_npreg_insample: ‘...’
  • Undocumented arguments in cycle_npreg_loglik: grids, method.type,
    method.grid ‘...’
  • Undocumented arguments in cycle_npreg_mstep: method.trend, polyorder,
    ncores, ‘...’
  • Undocumented arguments in cycle_npreg_outsample: ‘...’
  • Undocumented arguments in initialize_grids: ‘...’
  • Undocumented arguments in run_seurat: s.genes, g2m.genes, n.bin,
    seed.use, random.seed
  • ALL FIXED!

• Please fix: Add an inst/CITATION file; see
  github.com/stephenslab/mixsqp for an example.

  • fixed!

• Suggestion: In the file circle_npreg.R, the functions are
  "cycle_npreg_*". Maybe hange the name of the file to be consistent.

  • changed to cycle_npreg.R!

• Suggestion: Explain in "Description" field of DESCRIPTION file why
  it is called "peco". (And in README!)

  • PrEdicting Cell cycle prOgression in a
    continuum using scRNA-seq data

• Suggestion: If reasonable to do so, reduce size of data sets in
  "data" folder in order to reduce installation time from GitHub.

• Suggestion: Create a "release" on GitHub. (I can show you how.)

  • First realease 0.6.1