#pragma GCC optimize("Ofast")
//#include "grader.cpp"
#include<bits/stdc++.h>
using namespace std;
#define pb push_back
#define fi first
#define se second
#define sz(a) (int)(a.size())
#define all(a) a.begin(),a.end()
#define lb lower_bound
#define ub upper_bound
#define owo ios_base::sync_with_stdio(0);cin.tie(0);
#define MOD (ll)(998244353)
#define INF (ll)(1e18)
#define debug(...) fprintf(stderr, __VA_ARGS__),fflush(stderr)
#define time__(d) for(long blockTime = 0; (blockTime == 0 ? (blockTime=clock()) != 0 : false);\
debug("%s time : %.4fs\n", d, (double)(clock() - blockTime) / CLOCKS_PER_SEC))
typedef long long int ll;
typedef long double ld;
typedef pair<ll,ll> PII;
typedef pair<int,int> pii;
typedef vector<vector<int>> vii;
typedef vector<vector<ll>> VII;
struct fenwick{
ll n;
vector<ll>bit;
void upd(ll pos,ll val){
for(int i=pos;i<=n;i+=i&(-i))bit[i]+=val;
}
ll query(ll r){
ll res = 0;
for(int i=r;i>0;i-=i&(-i))res+=bit[i];
return res;
}
void init(vector<ll>a){
for(int i=1;i<=n;i++){
bit[i]+=a[i];
int j = i + (i&(-i));
if(j<=n)bit[j]+=bit[i];
}
}
int search(ll val){
ll sum = 0,pos = 0;
for(int i=19;i>=0;i--){
if(pos+(1<<i) < n && sum+bit[pos+(1<<i)]<val){
pos+=(1<<i);
sum+=bit[pos];
}
}
return pos+1;
}
fenwick(int _n){
n = _n;
bit.resize(n+1);
}
};
long long count_swaps(std::vector<int> s) {
int n = sz(s)/2;
vii l(n+1),r(n+1);
for(int i=0;i<2*n;i++){
if(s[i] < 0)l[-s[i]].pb(i);
else r[s[i]].pb(i);
}
vector<int>pos(2*n+1);
vector<pair<int,pii>>shoe;
for(int i=1;i<=n;i++){
for(int j=0;j<sz(l[i]);j++){
shoe.pb({l[i][j] + r[i][j],{l[i][j],r[i][j]}});
}
}
sort(all(shoe));
int cnt = 1;
for(auto x:shoe){
int a = x.se.fi;
int b = x.se.se;
pos[a] = cnt;
pos[b] = cnt+1;
cnt+=2;
}
ll ans = 0;
fenwick val(2*n+1);
for(int i=0;i<2*n;i++){
ans+=val.query(2*n) - val.query(pos[i]);
val.upd(pos[i],1);
}
return ans;
}
