Update datatable-intro.Rmd

[gurbuxanink]

A powerful feature of data.table is the ability to store lists as elements (a facility not available in data.frame). Here is a simple illustration of storing a fitted model as data.table element.

When statistical models are fit to different groups within a data.table, these models can be stored in a result data.table. Now this resulting data.table can be used to access fitted models.

[tdeenes]

note that this is not exactly true, data.frames can hold list elements:

datfr <- data.frame(idx = 1:3)
datfr$lst <- lapply(datfr$idx, seq_len)
str(datfr)
# 'data.frame':	3 obs. of  2 variables:
#  $ idx: int  1 2 3
#  $ lst:List of 3
#   ..$ : int 1
#   ..$ : int  1 2
#   ..$ : int  1 2 3

Having said that, of course it is much more convenient to use data.tables for working with this kind of objects. E.g.:

lst <- lapply(1:3, seq_len)
## this works
data.table(idx = seq_along(lst), lst = lst)
## this fails
data.frame(idx = seq_along(lst), lst = lst)

[tdhock]

please clarify your proposed addition by mentioning that it is possible in frame but requires a bit more typing, since you have to wrap lists with I()

> data.table(L=list(1:2))
        L
   <list>
1:    1,2
> data.frame(L=list(1:2))
  X1.2
1    1
2    2
> data.frame(L=I(list(1:2)))
     L
1 1, 2

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.62%. Comparing base (bb9faf6) to head (03de9eb).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #6558   +/-   ##
=======================================
  Coverage   98.62%   98.62%           
=======================================
  Files          79       79           
  Lines       14450    14450           
=======================================
  Hits        14251    14251           
  Misses        199      199           

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

[tdhock]

that looks better thanks