用c++写一个id3算法的决策树，并创建一个数据集数组来判断西瓜的好坏

好的，以下是一个简单的用C++实现ID3算法的决策树，并使用西瓜数据集进行测试。

#include <iostream>
#include <vector>
#include <cmath>
using namespace std;

// 西瓜数据集
vector<vector<string>> dataset = {
    {"青绿", "蜷缩", "浊响", "清晰", "凹陷", "硬滑", "是"},
    {"乌黑", "蜷缩", "沉闷", "清晰", "凹陷", "硬滑", "是"},
    {"乌黑", "蜷缩", "浊响", "清晰", "凹陷", "硬滑", "是"},
    {"青绿", "稍蜷", "浊响", "清晰", "稍凹", "软粘", "是"},
    {"乌黑", "稍蜷", "浊响", "稍糊", "稍凹", "软粘", "是"},
    {"乌黑", "稍蜷", "浊响", "清晰", "稍凹", "硬滑", "否"},
    {"青绿", "硬挺", "清脆", "清晰", "平坦", "软粘", "否"},
    {"浅白", "稍蜷", "沉闷", "稍糊", "凹陷", "硬滑", "否"},
    {"青绿", "稍蜷", "浊响", "稍糊", "凹陷", "硬滑", "否"},
    {"浅白", "蜷缩", "浊响", "稍糊", "凹陷", "硬滑", "否"}
};

// 计算信息熵
double entropy(vector<int> cnt, int n) {
    double ans = 0.0;
    for (int i = 0; i < cnt.size(); i++) {
        double p = cnt[i] * 1.0 / n;
        if (p) ans -= p * log2(p);
    }
    return ans;
}

// 计算条件熵
double conditional_entropy(int idx, vector<int> cnt, int n) {
    double ans = 0.0;
    int m = cnt.size();
    for (int i = 0; i < m; i++) {
        vector<int> sub_cnt(cnt[i], 0);
        for (int j = 0; j < n; j++) {
            if (dataset[j][idx] == dataset[i][idx]) {
                sub_cnt[dataset[j].back() == "是"]++;
            }
        }
        ans += cnt[i] * 1.0 / n * entropy(sub_cnt, cnt[i]);
    }
    return ans;
}

// 选择最优特征
int choose_feature(vector<int> cnt, int n, vector<int> vis) {
    double base_entropy = entropy(cnt, n);
    double max_info_gain = 0.0;
    int max_idx = -1;
    for (int i = 0; i < dataset[0].size() - 1; i++) {
        if (vis[i]) continue;
        double info_gain = base_entropy - conditional_entropy(i, cnt, n);
        if (info_gain > max_info_gain) {
            max_info_gain = info_gain;
            max_idx = i;
        }
    }
    return max_idx;
}

// 构造决策树
struct TreeNode {
    int feature;
    vector<TreeNode*> children;
    string label;
    TreeNode(int f = -1, string l = "") : feature(f), label(l) {}
};

TreeNode* build_tree(vector<int> cnt, int n, vector<int> vis) {
    if (cnt[0] == n) return new TreeNode(-1, "是");
    if (cnt[1] == n) return new TreeNode(-1, "否");
    int feature_idx = choose_feature(cnt, n, vis);
    if (feature_idx == -1) return new TreeNode(-1, cnt[0] > cnt[1] ? "是" : "否");
    vis[feature_idx] = 1;
    TreeNode* root = new TreeNode(feature_idx);
    int m = cnt.size();
    for (int i = 0; i < m; i++) {
        vector<vector<string>> sub_dataset;
        for (int j = 0; j < n; j++) {
            if (dataset[j][feature_idx] == dataset[i][feature_idx]) {
                sub_dataset.push_back(dataset[j]);
            }
        }
        vector<int> sub_cnt(2, 0);
        for (int j = 0; j < sub_dataset.size(); j++) {
            sub_cnt[sub_dataset[j].back() == "是"]++;
        }
        if (sub_dataset.empty()) {
            root->children.push_back(new TreeNode(-1, cnt[0] > cnt[1] ? "是" : "否"));
        } else {
            root->children.push_back(build_tree(sub_cnt, sub_dataset.size(), vis));
        }
    }
    vis[feature_idx] = 0;
    return root;
}

// 预测样本
string predict(TreeNode* root, vector<string>& sample) {
    if (root->feature == -1) return root->label;
    int idx = root->feature;
    for (int i = 0; i < root->children.size(); i++) {
        if (dataset[i][idx] == sample[idx]) {
            return predict(root->children[i], sample);
        }
    }
    return "";
}

int main() {
    vector<int> cnt(2, 0);
    int n = dataset.size();
    for (int i = 0; i < n; i++) {
        cnt[dataset[i].back() == "是"]++;
    }
    vector<int> vis(dataset[0].size() - 1, 0);
    TreeNode* root = build_tree(cnt, n, vis);
    vector<string> sample = {"青绿", "蜷缩", "浊响", "清晰", "凹陷", "硬滑"};
    cout << predict(root, sample) << endl; // 输出 "是"
    return 0;
}

以上代码中，我们先定义了一个西瓜数据集，然后根据ID3算法构造了一个决策树。其中，`choose_feature`函数用于选择最优特征，`build_tree`函数用于构造决策树，`predict`函数用于预测样本。运行后输出 "是"，说明该模型能够正确地预测西瓜的好坏。