-
-
Save elvircrn/79e694c4d9b446f16546305865f659d8 to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#include <iostream> | |
#include <cstdio> | |
#include <cstdlib> | |
#include <queue> | |
#include <cmath> | |
#include <stack> | |
#include <map> | |
#include <set> | |
#include <algorithm> | |
#include <sstream> | |
#include <unordered_map> | |
#include <functional> | |
#include <utility> | |
namespace patch | |
{ | |
template<class T> std::string to_string(const T &n) | |
{ | |
std::ostringstream stm; | |
stm << n; | |
return stm.str(); | |
} | |
} | |
template<class T> | |
class HuffmanTree | |
{ | |
private: | |
typedef std::pair<T, int> Node; | |
typedef std::pair<double, T> Data; | |
struct Edge | |
{ | |
T v; | |
char c; | |
Edge() { } | |
Edge(const Node &_v, char _c) : v(_v), c(_c) { } | |
}; | |
std::unordered_map<T, T> parent; | |
std::unordered_map<T, double> prob; | |
std::map<Node, std::vector<Node>> graph; | |
std::vector<char> codes; | |
std::unordered_map<T, int> groupSize; | |
Node root; | |
std::unordered_map<T, std::string> huffman; | |
int TakeCount(int n) const | |
{ | |
return 2 + ((n - 4) % (codes.size() - 1)); | |
} | |
void InitValues(const std::vector<std::pair<T, double>> &dataSet) | |
{ | |
for (auto& x : dataSet) | |
{ | |
prob[x.first] = x.second; | |
parent[x.first] = x.first; | |
groupSize[x.first] = 1; | |
} | |
} | |
T Find(const T &x) | |
{ | |
if (parent[x] == x) | |
return x; | |
else | |
return parent[x] = Find(parent[x]); | |
} | |
Node FindNode(const T &x) | |
{ | |
return Node(x, groupSize[x]); | |
} | |
Node Union(const T &x, const T &y) | |
{ | |
T fX = Find(x); | |
T fY = Find(y); | |
if (groupSize[fX] < groupSize[fY]) | |
{ | |
groupSize[fY] += groupSize[fX]; | |
parent [fX] = fY; | |
return Node(fY, groupSize[fY]); | |
} | |
else | |
{ | |
groupSize[fX] += groupSize[fY]; | |
parent [fY] = fX; | |
return Node(fX, groupSize[fX]); | |
} | |
} | |
void Huffman() | |
{ | |
std::priority_queue<Data, | |
std::vector<Data>, | |
std::function<bool(const Data &a, const Data &b)>> pq ( | |
[](const Data &a, const Data &b) -> bool | |
{ | |
return a.first > b.first; | |
}); | |
for (auto& x : prob) | |
pq.push(Data(x.second, x.first)); | |
Node node; | |
while (pq.size() > 1) | |
{ | |
int m = TakeCount((int)pq.size()); | |
auto mainElem = pq.top(); | |
pq.pop(); | |
Node lastNode = FindNode(mainElem.second); | |
double count = mainElem.first; | |
std::vector<Node> nodes(m - 1); // TODO: Optimize | |
for (int i = 0; i < m - 1; i++) | |
{ | |
auto help = pq.top(); | |
pq.pop(); | |
count += help.first; | |
nodes [i] = FindNode(help.second); | |
node = Union(help.second, mainElem.second); | |
} | |
nodes.push_back(lastNode); | |
for (Node& n : nodes) | |
graph[node].push_back(n); | |
pq.push(Data(count, node.first)); | |
} | |
root = node; | |
} | |
void Traverse(const Node ¤t, std::string buffer, bool write = false) | |
{ | |
int index = 0; | |
for (auto& next : graph[current]) | |
Traverse(next, buffer + codes[index++]); | |
if (!graph[current].size()) | |
{ | |
huffman[current.first] = buffer; | |
if (write) | |
std::cout << current.first << ' ' << buffer << '\n'; | |
} | |
} | |
public: | |
static std::vector<std::pair<T, double>> GetProb(const std::vector<std::pair<T, int>> &in) | |
{ | |
std::vector<std::pair<T, double>> v; | |
int total = 0; | |
for (auto &x : in) | |
total += x.second; | |
for (auto &x : in) | |
v.emplace_back(x.first, (double)x.second / total); | |
return v; | |
} | |
HuffmanTree(const std::vector<std::pair<T, double>> &dataSet, | |
const std::vector<char> &_codes) : codes(_codes) | |
{ | |
InitValues(dataSet); | |
Huffman(); | |
} | |
void Join() | |
{ | |
Traverse(root, ""); | |
} | |
void Analyse() | |
{ | |
std::unordered_map<T, int> n; | |
for (auto &e : huffman) | |
n[e.first] = e.second.length(); | |
double nsr = 0.0; | |
for (auto &e : huffman) | |
nsr += prob[e.first] * n[e.first]; | |
std::cout << "nsr = " << nsr << '\n'; | |
double hinf = 0.0; | |
for (auto &e : huffman) | |
hinf -= prob[e.first] * std::log2(prob[e.first]); | |
std::cout << "hinf = " << hinf << '\n'; | |
std::cout << "brzina prenosa = " << hinf / nsr << '\n'; | |
std::cout << "procenat iskoristenosti = " << (hinf / nsr) * 100.0 << "%\n"; | |
} | |
}; | |
int main() | |
{ | |
HuffmanTree<std::string> test ({{ "x3", 0.25 }, | |
{ "x6", 0.25 }, | |
{ "x1", 0.2 }, | |
{ "x5", 0.15 }, | |
{ "x4", 0.1 }, | |
{ "x2", 0.05 } | |
}, { '0', '1' }); | |
test.Join(); | |
test.Analyse(); | |
return 0; | |
} | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment