#include <bits/stdc++.h>
#pragma GCC optimize("Ofast,unroll-loops")
using namespace std;
const int N = 1e5 + 10, M = 100 + 10, inf = 1e9 + 10;
int n, k, a[N], dp[N][M], dp2[N];
struct node {
int mini, lazy;
} seg[N << 2];
void add(int l1, int r1, int l2, int r2, int v, int id) {
if (l1 >= r2 || r1 <= l2) return;
if (l1 >= l2 && r1 <= r2) {
seg[id].mini += v, seg[id].lazy += v;
return;
}
int mid = (l1 + r1) / 2;
add(l1, mid, l2, r2, v, id << 1), add(mid, r1, l2, r2, v, id << 1 | 1);
seg[id].mini = min(seg[id << 1].mini, seg[id << 1 | 1].mini) + seg[id].lazy;
}
void input() {
cin >> n >> k;
for (int i = 0; i < n; i++)
cin >> a[i];
}
void solve() {
int tmp = 0;
for (int i = 0; i < n; i++) {
dp2[i] = i - 1;
while (dp2[i] > -1 && a[dp2[i]] <= a[i])
dp2[i] = dp2[dp2[i]];
tmp = max(tmp, a[i]);
dp[i][1] = tmp;
}
for (int j = 2; j <= k; j++) {
for (int i = 0; i < (n << 2); i++)
seg[i] = {inf, 0};
for (int i = 0; i < n; i++) {
if (i > 0)
add(0, n, i - 1, i, a[i] + dp[i - 1][j - 1] - inf, 1);
int p = i - 1;
while (p > dp[i]) {
add(0, n, dp2[p], p, a[i] - a[p], 1);
p = dp2[p];
}
dp[i][j] = seg[1].mini;
}
}
cout << dp[n - 1][k];
}
int main() {
ios:: sync_with_stdio(0), cin.tie(0), cout.tie(0);
input();
solve();
}
