1

I have several cases where I am trying to save a vector of elements to my global environment to be able to use for querying a database or other subsequent analyses.

As an example, I have a dataframe below, but in reality my data consists of thousands of values.

df:
example start   stop   range
set1    1       4180   1:4180
set2    460     499    460:499
set3    1       1127   1:1127
set4    1128    1622   1128:1622
set5    3741    3748   3741:3748

With this dataframe I want to capture every integer between start and stop, so I would like to use the range column to create a vector of every integer.

I have been using: 

cat(paste0(df$range, collapse = ", "))

Which then prints to the console:

1:4180, 460:499, 1:1127, 1128:1622, 3741:3748

I have then been copy/pasting that output to save it as a vector in my global environment:

nums <- c(1:4180, 460:499, 1:1127, 1128:1622, 3741:3748)

Once I have this vector I can then do additional analyses such as:

> length(nums)
[1] 5850
> length(unique(nums))
[1] 4180
> summary(nums)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
      1     712    1443    1727    2726    4180

etc. etc.

Is there a simple way to skip the copy/paste step and save my desired formatting as a vector directly?

kholden
  • 15
  • 3

1 Answers1

1

An option is Map to get the sequence (:) between the corresponding 'start', 'stop' values and unlist the list of vectors to a single vector. 

vec <- unlist(Map(`:`, df$start, df$stop))

Or if we want to use the 'range' column, then evaluate it

vec1 <- sapply(df$range, function(x) eval(parse(text = x)))
akrun
  • 874,273
  • 37
  • 540
  • 662
  • Thanks, that first option works great. I'm having issues with the second option because it's pulling the factor index from the range column, rather than the text itself and results in `> length(unique(vec1)) [1] 5` – kholden Apr 14 '20 at 18:41
  • @kholden in that case, just change it to `eval(parse(text = as.character(x)))` – akrun Apr 14 '20 at 18:42
  • 1
    I also had to add an unlist() in there, but now both options work well, thanks so much! – kholden Apr 14 '20 at 18:56
  • @kholden yes, `sapply` returns a vector only if the length are same in the `list` elements. For that reason, I always use `lapply` and then `unlist` (at least the behavior can be predetermined) – akrun Apr 14 '20 at 18:58