Segmentation fault, while running a Monte-Carlo-random-walk simulation in c++ - c++

Question

I am trying to simulate a random walk of 2000 particles, while the one boundary has the ability to make particles bound on that and merely perform a biased step.
There are of course probabilities for binding unbinding etc...
Below I have the whole code.
However I get segfault error.
I put some print statements in the code to see where the issue lies. But nothing. What I found strange though, is that although seed is fixed, the length of the output statement determined the loop, where code crushed.
I am totally inexperienced in these issues, so if you have any idea on what I could do, would be appreciated.
#include <iostream>
#include <stdio.h>
#include <stdlib.h>
#include <fstream>
#include <string>
using namespace std;
const int pi=6;
const int epsilon=10;
const int X=3000;
const int Y=30; //length
const int time_steps=100000;
const int N=2000; // number of molecules
int kinesins[N][3]={0};//[X,Y,bound or not]
int grid[X][Y][2]={0};
void place_kinesins(){
for (int i=0; i<N;i++){
int x= rand()%X;
int y= (rand()%(Y-2))+2;
if (grid[x][y][0]==0){
kinesins[i][0]=x;
kinesins[i][1]=y;
kinesins[i][2]=0;
grid[x][y][0]=1;
}else{i--;}
}
}
void create_boundaries(){
for(int i=0;i<Y;i++){
grid[0][i][1]=-1;
grid[X-1][i][1]=-3;
}
for (int i=0; i<X; i++){
grid[i][Y-1][1]=-2;
}
}
void create_filament(){ //in order to create binding affinity.
for(int i=0; i<X;i++){
grid[i][1][1]=pi;
}
}
void step(int kinesin, int x_step, int y_step){
int x=kinesins[kinesin][0];
int y=kinesins[kinesin][1];
int x_end=x+x_step;
int y_end=y+y_step;
if (grid[x_end][y_end][0]==0){
grid[x][y][0]=0;
kinesins[kinesin][0]=x_end;
kinesins[kinesin][1]=y_end;
grid[x_end][y_end][0]=1;
}
}
void bound(int kinesin){
int roll=rand()%10000 ;
if (roll<epsilon){
kinesins[kinesin][2]=0;
step(kinesin,0,1);
}else{
if (roll%63==0){ //controls the binding rate speed
step(kinesin, 1,0);
};
}
}
void unbound(int kinesin){
cout<<"1";
int x= kinesins[kinesin][0];
int y= kinesins[kinesin][1];
int type= grid[x][y][1];
switch(type){
case 0:{
cout<<"2";
int roll=rand()%4;
switch(roll){
case 0:
step(kinesin,-1,0);
break;
case 1:
step(kinesin,1,0);
break;
case 2:
step(kinesin,0,1);
break;
case 3:
step(kinesin,0,-1);
break;
}
break;
}
case -1:
step(kinesin,1,0);
break;
case -2:
step(kinesin,0,-1);
break;
case -3:
step(kinesin,-1,0);
break;
default:
int roll=rand()%10000;
if(roll<grid[x][y][1]){kinesins[kinesin][2]=1;}
else{ if(roll%2==0){step(kinesin,0,1);}}
}
}
void kinesin_move(int kinesin){
cout<<" "<<kinesins[kinesin][0]<<kinesins[kinesin][1];
if (kinesins[kinesin][2]==0){
unbound(kinesin);
}else{
cout<<"3";
bound(kinesin);
}
}
void simulation(){
for(int j=7000; j<time_steps;j++){
cout<<endl<< j<<" "<<endl;
for (int kin=0; kin<N; kin++){
cout<<kin;
kinesin_move(kin);
cout<<"E " ;
}
}
}
void programm(){
srand(1);
create_boundaries();
create_filament();
cout<<"Filament done"<<endl;
place_kinesins();
cout<<"Kines placed"<<endl;
simulation();
}
int main(){
programm();
return 0;
}

Answer 1

Problem:
In the function step you're accessing the array grid out of its bonds, which produces Undefined Behaviour and the segmentation fault.
This can be proven adding an assert before if (grid[x_end][y_end][0]==0):
if(!(x_end < X && x_end >= 0))
std::cerr << x_end << std::endl;
assert(x_end < X && x_end >= 0);
if (grid[x_end][y_end][0]==0){
grid[x][y][0]=0;
kinesins[kinesin][0]=x_end;
kinesins[kinesin][1]=y_end;
grid[x_end][y_end][0]=1;
}
Output:
3000
Assertion failed: x_end < X && x_end >= 0 main.cpp line 57
This application has requested the Runtime to terminate it in an unusual way.
Solution:
You will have to check the arguments for step won't make it go out of bonds before each call.
Additional information:
using namespace std; is considered a bad practice (More info here).
rand is not uniformly distributed. You may want use the <random> library instead (More info here).
Global variables are bad (More info here).
You may want to use std::array instead of raw, C-style arrays.