D.cpp

//#pragma GCC optimize "trapv"
#pragma GCC optimize("Ofast")
#pragma GCC optimize("O3", "unroll-loops")
#pragma GCC target("avx,avx2,fma")
#include<bits/stdc++.h>
#define ll long long int
#define fab(a,b,i) for(int i=a;i<b;i++)
#define pb push_back
#define db double
#define mp make_pair
//#define endl "\n"
#define f first
#define se second
#define all(x) x.begin(),x.end()
#define MOD 1000000007
#define vll vector<ll>
#define vi vector<int>
#define pii pair<int,int>
#define pll pair<ll,ll>
#define quick ios_base::sync_with_stdio(false);cin.tie(NULL);cout.tie(NULL)
ll add(ll x, ll y) {ll res = x + y; return (res >= MOD ? res - MOD : res);}
ll mul(ll x, ll y) {ll res = x * y; return (res >= MOD ? res % MOD : res);}
ll sub(ll x, ll y) {ll res = x - y; return (res < 0 ? res + MOD : res);}
ll power(ll x, ll y) {ll res = 1; x %= MOD; while (y) {if (y & 1)res = mul(res, x); y >>= 1; x = mul(x, x);} return res;}
ll mod_inv(ll x) {return power(x, MOD - 2);}

using namespace std;

int main()
{	quick;
#ifndef ONLINE_JUDGE
	freopen("D:/sublime/input.txt", "r", stdin);
	freopen("D:/sublime/output.txt", "w", stdout);
#endif
	int t;
	cin >> t;
	while (t--)
	{
		int n;
		cin >> n;
		vector<int> a(n);
		fab(0, n, i)
		cin >> a[i];
		vector<int> pref(n + 1);
		pref[0] = a[0];
		fab(1, n, i)
		{
			pref[i] += pref[i - 1] + a[i];
		}
		int ans = n - 1;
		fab(0, n , i)
		{
			int rq = pref[i];
			int j = i + 1;
			bool ok = 1;
			//cout << "I:" << i << endl;
			int val = i;
			while (j < n)
			{
				int tot = 0;
				int prev = j;
				//cout << "jj:" << j << endl;
				while (j < n and tot < rq)
				{
					tot += a[j];
					j++;
				}
				if (tot == rq)
				{
					val += (j - prev - 1);

				}
				else if (tot != rq)
				{
					ok = 0;
					break;

				}

				//cout << "j:" << j << endl;

			}
			if (ok)
			{
				//cout << "i:" << i << endl;
				ans = min(ans, val);
			}
		}
		cout << ans << endl;
	}


	cerr << "time taken : " << (float)clock() / CLOCKS_PER_SEC << " secs" << endl;
	return 0;
}