Killing a random directed link in both direction in netlogo - network-programming

Question

Greeting,
I am very new to netlogo, and is very interested in GIS and network feature in netlogo.
Currently, I have constructed a directed graph network.
I would like to implement a function to kill a directed link path in both direction.
I have found way to kill a link in both direction if you know which are the linkage of the 2 node to kill off. Below is the code for that (by Seth Tisue in one of the question thread) :
ask a [ ask link-with b [ die ] ]
I would like to kill of a random linkage in both direction. I assume I would need to use "ask one-of links [ die ]" and another part of the code to kill of the corresponding other direction of the linkage.
I guess my question would be how the mechanism of "one-of" work so that I would be able to store this random link first, and then kill off both direction using the first code above.
As a side line, I am also puzzled by how netlogo work with variable which is quite unlike normal C++ program(my fundamental is with c++ coding). I do quite a lot of importing for my network and GIS shapefile, and currently I am still not using netlogo engine as my computational engine(a waste for me).
How do you store a simple "x=1+1" into a global variable in netlogo?
Really sorry if this question is too simple for most.

Answer 1

Just as a general stackoverflow convention, if you have two questions, please ask them as separate questions. This is to make stackoverflow a useful resource for programmers, so that the title and question and answer are both together and easily found.
First, how do you even know there is a link in both directions? Typically directed graphs have many pairs of nodes where there is A to B or B to A but not both. The code below tests the other direction and kills both.
I have included your example of setting variable x to the value of 1 + 1 in the setup.
I am convinced there is a better way of doing this as NetLogo code is usually less ugly, but I can't find it. The following code is a complete runnable example that creates a bunch of nodes and links and then kills a link (including the reciprocal if it exists) whenever the kill-link procedure is run.
globals [x tonode fromnode]
to setup
clear-all
create-turtles 20
[ setxy random-xcor random-ycor]
repeat 100
[ ask one-of turtles
[ create-link-to one-of other turtles ; will not be created if already link
]
]
set x 1 + 1
end
to kill-link
ask one-of links
[ set fromnode end1
set tonode end2
die
]
ask tonode
[ if out-link-neighbor? fromnode
[ ask out-link-to fromnode [die]
]
]
end

Answer 2

Global variable should be declared at the beginning of program:
globals [
my-global-link; this will be my global variable
my-another-one-link; another one
]
To "store" random link use it like:
to do-something
set my-global-link one-of links
end
For local variables (inside a procedure) just use let do declare it first:
to do-something
let tmp-global-link one-of links
end