#include <iostream>
#include <vector>
#include <algorithm>
#include <map>
using namespace std;
const long long INF = 4e18; // Using a large enough value for infinity
vector<int> adj[100005];
int emp_lang[100005];
int level[100005];
int subtree_sz[100005];
int n_nodes, k_langs;
long long max_p_global = 0;
long long min_s_global = INF;
// Sparse representation of C[k][l] (total employees of lang k at level l)
// relevant_Cs_at_level[l] stores a list of pairs {lang_k, count},
// where count = C[lang_k][l] and count > 0.
vector<vector<pair<int, int>>> relevant_Cs_at_level;
// Global arrays for Sack states
// current_P_val[k_test] = sum_l min(freq_level[l], C[k_test][l]) for current Sack bag
// current_I_val[lang_x] = count of employees with lang_x in current Sack bag
long long current_P_val[100005];
long long current_I_val[100005];
// freq_level[l] = count of employees at level l in current Sack bag
int freq_level[100005];
void dfs_levels(int u, int p, int l) {
level[u] = l;
for (int v : adj[u]) {
if (v == p) continue;
dfs_levels(v, u, l + 1);
}
}
void dfs_size(int u, int p) {
subtree_sz[u] = 1;
for (int v : adj[u]) {
if (v == p) continue;
dfs_size(v, u);
subtree_sz[u] += subtree_sz[v];
}
}
// Adds/removes node_x effects from global Sack state variables
void update_sack_data(int node_x, int val_sign) {
int lvl_x = level[node_x];
int lang_x = emp_lang[node_x];
long long old_freq_at_lvl_x = freq_level[lvl_x];
freq_level[lvl_x] += val_sign;
long long new_freq_at_lvl_x = freq_level[lvl_x];
// Update current_P_val[k_test] for all k_test based on the change of freq_level[lvl_x]
// Only languages k_test for which C[k_test][lvl_x] > 0 will have their min term potentially change.
if (lvl_x < (int)relevant_Cs_at_level.size()) { // Boundary check for lvl_x
for (auto const& pair_kc : relevant_Cs_at_level[lvl_x]) {
int k_test = pair_kc.first;
long long C_val_k_test_lvl_x = pair_kc.second;
long long term_change = 0;
// Contribution from min(old_freq, C_val)
if (old_freq_at_lvl_x > 0) {
term_change -= min(old_freq_at_lvl_x, C_val_k_test_lvl_x);
}
// Contribution from min(new_freq, C_val)
if (new_freq_at_lvl_x > 0) {
term_change += min(new_freq_at_lvl_x, C_val_k_test_lvl_x);
}
// Apply the net change to current_P_val for this k_test
current_P_val[k_test] += term_change;
}
}
// Update current_I_val for the language of the node being processed
current_I_val[lang_x] += val_sign;
}
// Helper to add/remove all nodes in u's subtree from Sack state
void dfs_process_subtree(int u, int p, int val_sign) {
update_sack_data(u, val_sign);
for (int v : adj[u]) {
if (v == p) continue;
dfs_process_subtree(v, u, val_sign);
}
}
// Main Sack algorithm DFS
void dfs_sack(int u, int p, bool keep) {
int max_child_sz = -1, big_child = -1;
for (int v : adj[u]) {
if (v == p) continue;
if (subtree_sz[v] > max_child_sz) {
max_child_sz = subtree_sz[v];
big_child = v;
}
}
// Process small children first, and clear their data from Sack state
for (int v : adj[u]) {
if (v == p || v == big_child) continue;
dfs_sack(v, u, false);
}
// Process big child (if any), keeping its data in Sack state
if (big_child != -1) {
dfs_sack(big_child, u, true);
// Sack state now reflects big_child's subtree
}
// Add node u itself to the Sack state
// (If there was a big child, its state is already there; if not, state is empty)
update_sack_data(u, 1);
// Add subtrees of small children to the Sack state
for (int v : adj[u]) {
if (v == p || v == big_child) continue;
dfs_process_subtree(v, u, 1);
}
// Now Sack state (freq_level, current_P_val, current_I_val) reflects subtree S_u
// Calculate P_u and S_u for u as team lead
int k_u = emp_lang[u]; // Project language is language of team lead u
long long cand_P = current_P_val[k_u];
long long cand_S = cand_P - current_I_val[k_u]; // Switches = Total Team - Initial Team
// Update global best
if (cand_P > max_p_global) {
max_p_global = cand_P;
min_s_global = cand_S;
} else if (cand_P == max_p_global) {
min_s_global = min(min_s_global, cand_S);
}
// If data for u's subtree is not to be kept (i.e., u is a small child or root of full Sack call)
if (!keep) {
// Remove all nodes of S_u (which includes u and all its descendants) from Sack state
dfs_process_subtree(u, p, -1);
}
}
int main() {
ios_base::sync_with_stdio(false);
cin.tie(NULL);
cin >> n_nodes >> k_langs;
for (int i = 0; i < n_nodes; ++i) {
cin >> emp_lang[i];
}
// Build directed adjacency list for the tree
vector<vector<int>> adj_build(n_nodes);
for (int i = 1; i < n_nodes; ++i) {
int b_val;
cin >> b_val;
adj_build[b_val].push_back(i);
}
for(int i=0; i<n_nodes; ++i) adj[i] = adj_build[i];
dfs_levels(0, -1, 0); // Compute levels for all nodes
// Precompute sparse C[k][l] into relevant_Cs_at_level
relevant_Cs_at_level.resize(n_nodes); // Max level is n_nodes-1
vector<map<int, int>> temp_counts_for_Cs(n_nodes);
for (int i = 0; i < n_nodes; ++i) {
if (level[i] < n_nodes) { // level[i] is already computed and < n_nodes
temp_counts_for_Cs[level[i]][emp_lang[i]]++;
}
}
for (int l = 0; l < n_nodes; ++l) {
if (!temp_counts_for_Cs[l].empty()) {
relevant_Cs_at_level[l].reserve(temp_counts_for_Cs[l].size());
for (auto const& pair_kl : temp_counts_for_Cs[l]) {
relevant_Cs_at_level[l].push_back({pair_kl.first, pair_kl.second});
}
}
}
// Initialize Sack global state arrays (static global arrays are zero-initialized by default in C++)
// Explicitly zeroing can be done for safety/clarity if needed, but not strictly necessary for global/static.
/*
for(int i=0; i < k_langs && i < 100005; ++i) {
current_P_val[i] = 0;
current_I_val[i] = 0;
}
for(int i=0; i < n_nodes && i < 100005; ++i) {
freq_level[i] = 0;
}
*/
dfs_size(0, -1); // Compute subtree sizes needed for Sack's big child heuristic
dfs_sack(0, -1, false); // Start Sack algorithm from root (node 0)
cout << max_p_global << " " << min_s_global << endl;
return 0;
}
# | Verdict | Execution time | Memory | Grader output |
---|
Fetching results... |
# | Verdict | Execution time | Memory | Grader output |
---|
Fetching results... |
# | Verdict | Execution time | Memory | Grader output |
---|
Fetching results... |
# | Verdict | Execution time | Memory | Grader output |
---|
Fetching results... |
# | Verdict | Execution time | Memory | Grader output |
---|
Fetching results... |