don't use head and tail … always be sampling from the middle of the data.frame

peek.R

taste <- function(soup, ladle=5L) sample.int(x=soup, size=ladle, replace=TRUE)

peek <- function(df,n=5L) df[ taste(nrow(df),n) ,   ]      #dataframe[r,c]  means "subset of dataframe row #r, column #c"

p <- function(df,n=5L)  rbind(head(df,1L), peek(df,n), tail(df,1L)) 


#SAMPLE OUTPUT
require(bigvis)
data(movies)
dim(movies)
#[1] 130456     14

#so movies is too long to look at.
#yet if I always peek at it with head(movies) I will get bored,
#and be wasting the opportunity to gradually get to know my dataset.


peek(movies)
#                                  title year length budget rating votes mpaa
#109844                      Liberty Kid 2007     92 200000    6.2   126 <NA>
#30142  Due mattacchioni al Moulin Rouge 1964     90     NA    3.7     9 <NA>
#87158                  Nunca es domingo 2002     19     NA    4.4    16 <NA>
#122636               Turning Point 1977 2009    110     NA    6.7     6 <NA>
#124824                         Hooters! 2010     90  62250    7.0     5 <NA>
#
#       Action Animation Comedy Drama Documentary Romance Short
#109844  FALSE     FALSE  FALSE FALSE       FALSE   FALSE FALSE
#30142   FALSE     FALSE  FALSE FALSE       FALSE   FALSE FALSE
#87158   FALSE     FALSE  FALSE FALSE       FALSE   FALSE FALSE
#122636  FALSE     FALSE  FALSE FALSE       FALSE   FALSE FALSE
#124824  FALSE     FALSE  FALSE FALSE       FALSE   FALSE FALSE





#TODO:
# - contiguous random pieces
# - deal with other shapes like lists