analyze function that collects various statistics

[jangorecki]

Built on top of lazy-forder branch, diff will include lazy-forder diff as well!
For useful diff see https://github.com/Rdatatable/data.table/compare/lazy-forder...analyze?expand=1
After merging lazy-forder will need rebase.
Closes #2879

x = data.table(id1=1:5, id2=c(NA,2:5), id3=rep(NA_integer_, 5L), id4=c("a","b","c","b","a"), id5=c("a","b","\u221A","d","e"), v1=1L/2, v2=c(NaN,2:5/2), v3=c(1:4/2, NaN), v4=c(-Inf,2:4,Inf))
cols = list("id1", "id2","id3", c("id1","id2"), "id4", "id5", c("id4","id5"), "v1", "v2", "v3", "v4")

a = analyze(x, cols, flat=TRUE)
a
#       cols ncols             coltype            colclass sorted uniqueN maxgrpN  anyNA anyInfNaN anyNotASCII anyNotUTF8  anyNF unique  const  allNA
#     <char> <int>              <char>              <char> <lgcl>   <int>   <int> <lgcl>    <lgcl>      <lgcl>     <lgcl> <lgcl> <lgcl> <lgcl> <lgcl>
# 1:     id1     1             integer             integer   TRUE       5       1  FALSE     FALSE       FALSE      FALSE  FALSE   TRUE  FALSE  FALSE
# 2:     id2     1             integer             integer   TRUE       5       1   TRUE     FALSE       FALSE      FALSE   TRUE   TRUE  FALSE  FALSE
# 3:     id3     1             integer             integer   TRUE       1       5   TRUE     FALSE       FALSE      FALSE   TRUE  FALSE   TRUE   TRUE
# 4: id1,id2     2     integer,integer     integer,integer   TRUE       5       1   TRUE     FALSE       FALSE      FALSE   TRUE   TRUE  FALSE  FALSE
# 5:     id4     1           character           character  FALSE       3       2  FALSE     FALSE       FALSE      FALSE  FALSE  FALSE  FALSE  FALSE
# 6:     id5     1           character           character  FALSE       5       1  FALSE     FALSE        TRUE      FALSE  FALSE   TRUE  FALSE  FALSE
# 7: id4,id5     2 character,character character,character  FALSE       5       1  FALSE     FALSE        TRUE      FALSE  FALSE   TRUE  FALSE  FALSE
# 8:      v1     1              double             numeric   TRUE       1       5  FALSE     FALSE       FALSE      FALSE  FALSE  FALSE   TRUE  FALSE
# 9:      v2     1              double             numeric   TRUE       5       1  FALSE      TRUE       FALSE      FALSE   TRUE   TRUE  FALSE  FALSE
#10:      v3     1              double             numeric  FALSE       5       1  FALSE      TRUE       FALSE      FALSE   TRUE   TRUE  FALSE  FALSE
#11:      v4     1              double             numeric   TRUE       5       1  FALSE      TRUE       FALSE      FALSE   TRUE   TRUE  FALSE  FALSE

a = analyze(x, cols) ## flat=FALSE returns first and last rows as nested DT

No tests yet. No manual yet. I will follow up with this after some initial feedback.

[codecov]

Codecov Report

Merging #4478 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master    #4478   +/-   ##
=======================================
  Coverage   99.42%   99.42%           
=======================================
  Files          72       72           
  Lines       14153    14153           
=======================================
  Hits        14071    14071           
  Misses         82       82           

---

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 9382cb8...9382cb8. Read the comment docs.