0

我无法弄清楚我的堆排序问题出在哪里。
该程序从命令行获取文件名，将单词导入向量，然后将该向量转换为向量对，vector<string,int>其中 string 是单词，int 是该单词在文件中的实例数。

然后vector<PAIR>按字符串（值或 v）或 int（键或 k）排序。我的按键排序工作正常，但按值排序已关闭。我怀疑我在按值排序时缺少 max_heapify 中的 if 语句。这是我的代码：

主文件

#include <fstream>
#include <iostream>
#include <stdlib.h>
#include <vector>
#include <string>
#include <string.h>
#include <stdio.h>
#include <map>
#include <time.h>
#include "readwords.h"

using namespace std;

readwords wordsinfile;
vector<string> allwords;
bool times;
char *filename;
timespec timestart,timeend;
vector< pair<string,int> > allwords_vp;

timespec diffclock(timespec start, timespec end);

int main ( int argc, char *argv[] ) {

    filename = argv[1];

    //Lets open the file 
    ifstream ourfile2(filename);

    //Lets get all the words using our requirements
    allwords = wordsinfile.getwords(ourfile2);
    //Convert all the words from file and count how many times they 
    //appear.  We will store them in a vector<string,int> string 
    //being the word and int being how many time the word appeared
    allwords_vp = wordsinfile.count_vector(allwords);

    cout << "HeapSort by Values" << endl;
        if (times) {
        clock_gettime(CLOCK_PROCESS_CPUTIME_ID, &timestart);
                wordsinfile.heapsort(const_cast<char *>("v"));
        clock_gettime(CLOCK_PROCESS_CPUTIME_ID, &timeend);
                cout << "HeapSort by Values ran in "
                     << diffclock(timestart,timeend).tv_nsec << " nanosecond or "
                     << diffclock(timestart,timeend).tv_nsec/1000 << " millisecond"
                     << endl;
        } else {
                wordsinfile.heapsort(const_cast<char *>("v"));
        }

    cout << "HeapSort by Keys" << endl;
        if (times) {
        clock_gettime(CLOCK_PROCESS_CPUTIME_ID, &timestart);
                wordsinfile.heapsort(const_cast<char *>("k"));
        clock_gettime(CLOCK_PROCESS_CPUTIME_ID, &timeend);
                cout << "HeapSort by Keys ran in "
                     << diffclock(timestart,timeend).tv_nsec << " nanosecond or "
                     << diffclock(timestart,timeend).tv_nsec/1000 << " millisecond"
                     << endl;
        } else {
                wordsinfile.heapsort(const_cast<char *>("k"));
        }
}

timespec diffclock(timespec start, timespec end) {
    timespec temp;
    if ((end.tv_nsec-start.tv_nsec)<0) {
        temp.tv_sec = end.tv_sec-start.tv_sec-1;
        temp.tv_nsec = 1000000000+end.tv_nsec-start.tv_nsec;
    } else {
        temp.tv_sec = end.tv_sec-start.tv_sec;
        temp.tv_nsec = end.tv_nsec-start.tv_nsec;
    }
    return temp;
}

读字.h

#ifndef READWORDS_H
#define READWORDS_H

#include <vector>
#include <map>
#include <utility>
#include <time.h>

typedef std::pair<std::string, int> PAIR;

bool isasciifile(std::istream& file);

class readwords {
    private:
         std::vector<PAIR> vp;
    public:
         std::vector<std::string> getwords(std::istream& file);
         std::vector<PAIR> count_vector(std::vector<std::string> sv);
         void print_vectorpair(std::vector<PAIR> vp);
         void print_vector(std::vector<std::string> sv);     
         void heapsort(char how[]);
         void buildmaxheap(std::vector<PAIR> &vp, int heapsize, char how[]);
         void max_heapify(std::vector<PAIR> &vp, int i, int heapsize, char how[]);
         void swap_pair(PAIR &p1, PAIR &p2);
};

读字.cpp

#include <fstream>
#include <iostream>
#include <map>
#include "readwords.h"
#include <vector>
#include <string>
#include <utility>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <time.h>

//using std::vector;
using namespace std;

typedef pair<string, int> PAIR;

// Do we have a ASCII file?
// Lets test the second 10 chars to make sure
// This method is flawed if the file is less than 10 chars
bool isasciifile(std::istream& file) {
    int c = 0;
    bool foundbin = false;
    for(c=0; c < 10;c++) {
        if(!isprint(file.get())){
            // Looks like we found a non ASCII file, or its empty.
            foundbin = true;
        }
    }
    return foundbin;
}

// This is our workhorse as it splits up the words based on our criteria and
// passes them back as a vector of strings.
vector<string> readwords::getwords(std::istream& file) {
    char c;
    string aword;
    vector<string> sv;

            //Let go through the file till the end  
                        while(file.good()) {
                                c = file.get();
                                if (isalnum(c)) {
                    //convert any uppercase to lowercase
                    if(isupper(c)) {
                                                c = (tolower(c));
                                        }
                    //if its a space lets go onto the next char
                                        if(isspace(c)) { continue; }
                    //everything looks good lets add the char to our word
                                        aword.insert(aword.end(),c);
                                } else {
                    //its not a alphnum or a space so lets skip it
                    if(!isspace(c)) { continue; }
                    //reset our string and increment
                                        if (aword != "") {sv.push_back(aword);}
                                        aword = "";
                                        continue;
                                }
                        }
    return sv;
}

vector<PAIR> readwords::count_vector(vector<string> sv) {
    unsigned int i = 0;
    int j = 0;
    int match = 0;

    // cout << "Working with these string: " << endl;
    // print_vector(sv);

    for (i=0; i < sv.size(); i++) {
        // cout << "count of i: " << i << " word is: " << sv.at(i) << endl;

        match = 0;
        if(readwords::vp.size() == 0) {
            readwords::vp.push_back(make_pair(sv.at(i),1)); continue;
        }

        for (j=readwords::vp.size() - 1; j >= 0; --j) {
            if (sv.at(i) == readwords::vp.at(j).first) {
                // cout << "Match found with: " << sv.at(i) << endl;;
                readwords::vp.at(j).second = readwords::vp.at(j).second + 1;
                match = 1;
            } 

            // cout << "Value of j and match: " << j << match << endl;
            if ( j == 0 && match == 0) {
                // cout << "Match found at end with: " << sv.at(i) << endl;;
                readwords::vp.push_back(make_pair(sv.at(i),1));
            }
        }
    }

    //Prob need to sort by first data type then second here, prior to sort functions.
    //Might not be the best place as the sort functions would alter it, if not here
    //then each sort requires to do secondary search

    return readwords::vp;
}

void readwords::print_vectorpair(vector<PAIR> vp) {
    unsigned int i = 0;

    for (i=0; i < vp.size(); ++i) {
        cout << vp.at(i).first << " " << vp.at(i).second << endl;
    }
}

void readwords::print_vector(vector<string> sv) {
    unsigned int i = 0;

    for (i=0; i < sv.size(); ++i) {
        cout << sv.at(i) << endl;
    }
}

void readwords::heapsort(char how[]) {
    int heapsize = (readwords::vp.size() - 1);

    buildmaxheap(readwords::vp, heapsize, how);

    for(int i=(readwords::vp.size() - 1); i >= 0; i--) {
        swap(readwords::vp[0],readwords::vp[i]);
        heapsize--;
        max_heapify(readwords::vp, 0, heapsize, how);
    }

    print_vectorpair(readwords::vp);
}

void readwords::buildmaxheap(vector<PAIR> &vp, int heapsize, char how[]) {

    for(int i=(heapsize/2); i >= 0 ; i--) {
        max_heapify(vp, i, heapsize, how);
    }
}

void readwords::max_heapify(vector<PAIR> &vp, int i, int heapsize, char how[]) {
    int left = ( 2 * i ) + 1;
    int right = left + 1;
    int largest;

    if(!strcmp(how,"v")) {  
        if(left <= heapsize && vp.at(left).second >= vp.at(i).second ) {
            if( vp.at(left).first >= vp.at(i).first ) {
                largest = left;
            } else {
                largest = i;
            }
        } else {
            largest = i;
        }

        if(right <= heapsize && vp.at(right).second >= vp.at(largest).second) {
            if( vp.at(right).first >= vp.at(largest).first) {
                largest = right;
            }
        }   
    }

    if(!strcmp(how,"k")) {  
        if(left <= heapsize && vp.at(left).first > vp.at(i).first) {
            largest = left;
        } else {
            largest = i;
        }

        if(right <= heapsize && vp.at(right).first > vp.at(largest).first) {
            largest = right;
        }   
    }

    if(largest != i) {
        swap(vp[i], vp[largest]);
        max_heapify(vp, largest, heapsize, how);
    }
}
4

2 回答 2

1

然后该向量按字符串（值或 v）或 int（键或 k）排序。

该描述与代码不匹配，仅使用第一个组件排序的how参数进行"k"排序，即string，并且使用"v"ashow参数进行排序将两个组件都考虑在内。

我认为通过 achar[]来确定排序标准是一个相当糟糕的主意，它应该是一个比较器函数，所以你只需要在max_heapify.

我的按键排序工作正常，但按值排序已关闭。我怀疑我在按值排序时缺少 max_heapify 中的 if 语句。

问题是堆排序需要总排序，否则无法正确排序。

你的条件

if(left <= heapsize && vp.at(left).second >= vp.at(i).second ) {
    if( vp.at(left).first >= vp.at(i).first ) {
        largest = left;
    } else {
        largest = i;
    }
} else {
    largest = i;
}

检查vp.at(left)(resp. right) 的两个组件是否至少与 的相应组件一样大vp.at(i)，从而导致产品偏序，两个通用对不具有可比性，在这种情况下，您max_heapify什么也不做。

例如，对于<"a",3>和<"b",2>在<"c",1>位置i, left, right中，无论以何种顺序，您都max_heapify设置largest为i。

如果您的排序依据"v"是int首先基于组件进行排序，并且在平局的情况下，请考虑string组件，您需要区分大小写vp.at(left).second > vp.at(i).second和相等性（right当然也是如此）。例如

if(left <= heapsize && vp.at(left).second >= vp.at(i).second ) {
    if(vp.at(left).second > vp.at(i).second || vp.at(left).first >= vp.at(i).first ) {
        largest = left;
    } else {
        largest = i;
    }
} else {
    largest = i;
}
于 2012-10-13T14:22:43.330 回答
-1

要按值排序 a vector<pair<string, int> >，请考虑添加vector<pair<int, string> >

vector<pair<int, string> > v(orignal.size());
for (int i = 0; i < v.size(); ++i) v[i] = make_pair(original[i].second, original[i].first);
sort(v.begin(), v.end());
于 2012-10-13T12:50:07.840 回答