NetLogo 3D: Printing multiple turtle variables to output at end of run and running through BehaviorSpace

I'm trying to find an efficient method of outputting a large number of turtle variables (20+) from a random selection of 100 turtles into an output field or text file. So far I have:

turtles-own [
variable1
variable2
variable3
variable4
.
.
.
]

to go 
if (ticks < 1) [reset-timer]
ticks
if count turtles >= end-population [
ask n-of 100 turtles [
output-show variable1
output-show variable2
output-show variable3
output-show variable4
]

I then get a list of the variables of each cell in a single column:

1

My question is how can I get these variable values on the same line so of the output or a text file so I can easily work with these data? Furthermore, how would I implement this in BehaviorSpace? Using this same command:

ask n-of 100 turtles [
output-show variable1
output-show variable2
output-show variable3
output-show variable4
]

... in the final commands field does not result in any of these data showing up in the output file.

Thanks!

There are a variety of ways to do this- for example, the csv extension is great candidate if you want to manually output your values. If you want to do this quickly in BehaviorSpace, here's how I usually go about it.

I'm assuming you want the variable values for the same 100 turtles each time, rather than sampling a new 100 turtles for each variable. So, I think the easiest way is to just make a globals variable for each of the variables of interest, then make a procedure to fill those lists as needed. For example, with this setup:

globals [ a-final b-final c-final ]

turtles-own [ a b c ]

to setup
  ca
  crt 100 
  reset-ticks
end

to go
  ask turtles [
    set a random 100
    set b one-of [ "Beep" "Boop" ]
    set c precision random-float 10 2
  ]
end

Each tick the turtles are just randomly updating their a, b, and c variables for the sake of this toy version. Then, you have a procedure that subsamples some number of turtles (here, 10) from your total population and updates the storage lists:

to output
  let selected-turtles n-of 10 turtles 
  set a-final [a] of selected-turtles
  set b-final [b] of selected-turtles
  set c-final [c] of selected-turtles
end

Now as long as that output runs right before your BehaviorSpace experiment ends, you can output those lists as a string, which you can easily separate and clean with R or similar software. For example, if you have a setup like:

You will get output that looks something like: