簡體   English   中英

我應該如何修復這些垃圾值?

[英]How should I fix these garbage values?

我是C語言的學徒,想在這里學習一些東西。 我遇到了一個問題,我希望我能自己解決它。 但是,我是一個矮小的包裝工,很容易拖延,所以我需要你的大力支持。 正如標題明確指出的那樣,我不知道為什么我的 topMatches() function 會吐出一些隨機浮點值,在 scores[] 數組的開頭和結尾打印出來。 像這樣,

Output #1
(-0.00000, Lisa Rose),(0.99124, Gene Seymour),(0.92447, Michael Phillips),(0.89341, Claudia Puig),(0.66285, Mick LaSalle),(0.38125, Jack Matthews),(-1.00000, Toby)
Output #2
(107185664961793568883398204719104.00000, Lisa Rose),(0.99124, Gene Seymour),(0.92447, Michael Phillips),(0.89341, Claudia Puig),(0.66285, Mick LaSalle),(0.38125, Jack Matthews),(-1.00000, Toby)
Output #3
(0.99124, Lisa Rose),(0.92447, Gene Seymour),(0.89341, Michael Phillips),(0.66285, Claudia Puig),(0.38125, Mick LaSalle),(-118195603315995709432961818167345152.00000, Jack Matthews),(-1.00000, Toby)
...

該值應在 -1 和 1 的范圍內。非常感謝您的反饋。

#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <math.h>

//data types
struct mInfo
{
    char mName[20];
    float rating;
};
struct cInfo
{
    char name[20];
    struct mInfo movi[7];
};
//prototype fxns
typedef double (*sim_fp)(struct cInfo *, const char *, const char *);
double sim_D(struct cInfo *prefs, const char *person1, const char *person2);
void topMatches(sim_fp fp, struct cInfo *prefs, const char *person1, int num);
int cmpFxn (const void * a, const void * b);
void reverseFxn(float arr[], int num);

int main() 
{
    int num = 7;
    struct cInfo critics[num];
  
    critics[0] = (struct cInfo) {"Lisa Rose", {"Lady in the Water", 2.5, "Snakes on a Plane", 3.5, "Just My Luck", 3, "Superman Returns", 3.5, "The Night Listener", 3, "You, Me and Dupree", 2.5}};
    
    critics[1] = (struct cInfo) {"Gene Seymour",{"Lady in the Water", 3, "Snakes on a Plane", 3.5, "Just My Luck", 1.5, "Superman Returns", 5, "The Night Listener", 3, "You, Me and Dupree", 3.5}};
    
    critics[2] = (struct cInfo) {"Michael Phillips",{"Lady in the Water", 2.5, "Snakes on a Plane", 3, "Superman Returns", 3.5, "The Night Listener", 4}};
    
    critics[3] = (struct cInfo) {"Claudia Puig",{"Snakes on a Plane", 3.5, "Just My Luck", 3, "Superman Returns", 4, "The Night Listener", 4.5, "You, Me and Dupree", 2.5}};
    
    critics[4] = (struct cInfo) {"Mick LaSalle",{"Lady in the Water", 3, "Snakes on a Plane", 4, "Just My Luck", 2, "Superman Returns", 3, "The Night Listener", 3, "You, Me and Dupree", 2}};
    
    critics[5] = (struct cInfo) {"Jack Matthews",{"Lady in the Water", 3, "Snakes on a Plane", 4, "Superman Returns", 5, "You, Me and Dupree", 3.5}};
    
    critics[6] = (struct cInfo) {"Toby",{"Snakes on a Plane", 4.5, "Superman Returns", 4, "You, Me and Dupree",1}};
    
    topMatches(sim_D, critics, "Toby", 7);

    return 0;
}

double sim_D(struct cInfo *prefs, const char *person1, const char *person2)
{
    int i=0;
    int x=0;
    int next=0;
    int p1;
    int p2;
    float X = 0;
    float Y = 0;
    float sumSq1 = 0;
    float sumSq2 = 0;
    float pSum = 0;
    float num = 0;
    float den = 0;
    float Pscore = 0;
    int nElements =0;
    
    
    for (i=0;i<7;i++) {
        if(strcmp(prefs[i].name, person1) ==0)
        {
           p1 = i; 
        }
        else if(strcmp(prefs[i].name, person2) ==0)
        {
            p2 = i;
        }
    }
  
    for (x=0;x<7;x++) {
        for (next=0;next<7;next++) 
        {
            if (!prefs[p1].movi[x].rating && !prefs[p2].movi[next].rating);
            else if (strcmp(prefs[p1].movi[x].mName, prefs[p2].movi[next].mName) == 0)
            {
                  X += prefs[p1].movi[x].rating;
                  Y += prefs[p2].movi[next].rating;
                  sumSq1 += pow(prefs[p1].movi[x].rating,2);
                  sumSq2 += pow(prefs[p2].movi[next].rating,2);
                  pSum += (prefs[p1].movi[x].rating*prefs[p2].movi[next].rating); 
                  nElements++;
            }
        }
        next = 0;
    }
  
    num = pSum-(X*Y/nElements);

    den=sqrt((sumSq1-pow(X,2)/nElements)*(sumSq2-pow(Y,2)/nElements));
    if(den ==0) return -1;

    Pscore = num/den; 

    return Pscore;
}

void topMatches(sim_fp fp, struct cInfo *prefs, const char *person1, int num)
{
    float scores[8];
    char *buf;
    int i=0;

    for (i=0;i<num;i++)
    {
        if(strcmp(person1, prefs[i].name)==0)
        {
          continue;
        }
        scores[i] = fp(prefs, person1, prefs[i].name);
    }
  
  qsort(scores, num, sizeof(float), (*cmpFxn));
  reverseFxn(scores, num);
   printf("\n\n");
        for(i=0;i<num;i++)
        {
          if (i == num-1)
          {
            printf("(%.5f, %s)", scores[num-1], prefs[num-1].name);
          }
          else
          {
            printf("(%.5f, %s),", scores[i], prefs[i].name);
          }
        }
}
void reverseFxn(float arr[], int num)
{
  float scoresTmp[num];
  int j;
        
    for(j=0;j<num;j++)
    {
      scoresTmp[num-1-j] = arr[j];
    }
      for(j=0;j<num;j++)
      {
        arr[j]=scoresTmp[j];
      }
}
int cmpFxn (const void * a, const void * b) 
{
   return ( *(int*)a - *(int*)b );
}

scores[] 數組的每個元素的值需要在 -1 到 1 的范圍內。

  1. 意外變量 output 提示正在使用未初始化的數據。 struct cInfo你有一個成員struct mInfo movi[7]; 但在main()中分配其中的 3 到 6 個。 通過從賦值切換到初始化,那些未顯式設置的記錄將被零初始化。 這似乎足以獲得一致的 output。
int main() {
    struct cInfo critics[] = {
        {"Lisa Rose", {
            {"Lady in the Water", 2.5},
            {"Snakes on a Plane", 3.5},
            {"Just My Luck", 3},
            {"Superman Returns", 3.5},
            {"The Night Listener", 3},
            {"You, Me and Dupree", 2.5}
        }},
        {"Gene Seymour", {
            {"Lady in the Water", 3},
            {"Snakes on a Plane", 3.5},
            {"Just My Luck", 1.5},
            {"Superman Returns", 5},
            {"The Night Listener", 3},
            {"You, Me and Dupree", 3.5}
        }},
        {"Michael Phillips", {
            {"Lady in the Water", 2.5},
            {"Snakes on a Plane", 3},
            {"Superman Returns", 3.5},
            {"The Night Listener", 4}
        }},
        {"Claudia Puig", {
            {"Snakes on a Plane", 3.5},
            {"Just My Luck", 3},
            {"Superman Returns", 4},
            {"The Night Listener", 4.5},
            {"You, Me and Dupree", 2.5}
        }},
        {"Mick LaSalle", {
            {"Lady in the Water", 3},
            {"Snakes on a Plane", 4},
            {"Just My Luck", 2},
            {"Superman Returns", 3},
            {"The Night Listener", 3},
            {"You, Me and Dupree", 2}
        }},
        {"Jack Matthews", {
            {"Lady in the Water", 3},
            {"Snakes on a Plane", 4},
            {"Superman Returns", 5},
            {"You, Me and Dupree", 3.5}
        }},
        {"Toby", {
            {"Snakes on a Plane", 4.5},
            {"Superman Returns", 4},
            {"You, Me and Dupree",1}
        }}
    };
    topMatches(sim_D, critics, "Toby", sizeof critics / sizeof *critics);
}

這是示例 output:

(0.99124, Lisa Rose),(0.92447, Gene Seymour),(0.89341, Michael Phillips),(0.66285, Claudia Puig),(0.38125, Mick LaSalle),(0.00000, Jack Matthews),(-1.00000, Toby)
  1. (不固定)考慮用空主體重寫不尋常的 if 表達式:
if (!prefs[p1].movi[x].rating && !prefs[p2].movi[next].rating);
else if (strcmp(prefs[p1].movi[x].mName, prefs[p2].movi[next].mName) == 0) {

到:

if (
   (prefs[p1].movi[x].rating || prefs[p2].movi[next].rating) && 
   !strcmp(prefs[p1].movi[x].mName, prefs[p2].movi[next].mName)
) {

如果您引入幾個虛榮變量,它可以寫得更緊湊。

  1. float替換為double並修復排序 function:
int cmpFxn (const void * a, const void * b) {
    double a2 = *(double *) a;
    double b2 = *(double *) b;
    if(a2 < b2) return -1;
    if(a2 > b2) return 1;
    return 0;
}
  1. valgrind 報告在 topMatches 中使用未初始化的數據,特別double scores[8]; . 您只迭代num ,即7 ,並且您沒有為對應於 person1 的條目設置 scores[i] 所以我建議您這樣做:
    double scores[num];
    memset(scores, 0, sizeof(scores));
  1. (不固定)檢查你的代碼是否有神奇的值,比如 7 和 20。如果你需要數組的大小,使用常量,傳入一個參數或sizeof a / sizeof *a (小心 arrays 在作為參數傳遞時降級為指針) .

至少這些問題:

錯誤比較 function

使用floatcmpFxn數組調用qsort() ,但cmpFxn()比較int *1

需要投射到float *2

// return (*(int*)a - *(int*)b )
return (*(float*)a > *(float*)b ) - (*(float*)a < *(float*)b );

不要使用return (*(float*)a - *(float*)b )因為float差異可能會溢出int范圍或截斷為 0。

代碼風險sqrt(some_negative)

注意輕微的浮點偽影,這些偽影會導致數學上永遠不應小於 0 的負數。

// den=sqrt((sumSq1-pow(X,2)/nElements)*(sumSq2-pow(Y,2)/nElements));
t = (sumSq1-pow(X,2)/nElements)*(sumSq2-pow(Y,2)/nElements);
den = t >= 0.0 ? sqrt(t) : 0;

floatdouble

代碼隨意混合floatdouble對象和函數。 建議只double

出於調試目的,很容易使用#define float double來臨時執行此操作。


*1
cmpFxn()也是int的弱比較 function,因為它有int溢出的風險。

// return ( *(int*)a - *(int*)b );
return (*(int*)a > *(int*)b ) - (*(int*)a < *(int*)b );

*2
如果涉及非數字,則比較更加復雜。 讓我們暫時假設情況並非如此。

暫無
暫無

聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.

 
粵ICP備18138465號  © 2020-2024 STACKOOM.COM