855 - Lunch in Grid City

[howardcheng]

I must be braindead. I cannot see why my program gives WA:

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

int cmp(const void *a, const void *b)
{
return *(int *)a - *(int *)b;
}

int main(void)
{
int T, F;
int s[50000], a[50000];
int i;
int meds, meda;

scanf("%d\n", &T);
while (T-- > 0) {
scanf("%*d %*d %d", &F);
for (i = 0; i < F; i++) {
scanf("%d %d", s+i, a+i);
}
qsort(s, F, sizeof(int), cmp);
qsort(a, F, sizeof(int), cmp);
if (F % 2) {
meds = s[F/2];
meda = a[F/2];
} else {
meds = (s[F/2-1]+s[F/2])/2;
meda = (a[F/2-1]+a[F/2])/2;
}
printf("(Street: %d, Avenue: %d)\n", meds, meda);
}
return 0;

}

[go]

I solved this problem in two ways but every time I got WA.Here is my source,my algorithm is dynamic,at first step I found the total distance from all points to (1,1) and then for every other points ,I find the distance
by the point just before it(it is a distance for last point+number of points in right side of the recent point - number of points left side the recent point) the distance with the recent point and points in it's right side decrease 1,in it's left side increase 1.
[cpp]
#include <iostream>
#include <string>
using namespace std;
#define int long long
int a,b,f,m[1001][1001],r[1001],l[1001],u[1001],d[1001];
int main()
{
int mm;
cin>>mm;
for(int tt=0;tt<mm;tt++)
{
cin>>a>>b>>f;
memset(m,0,sizeof m);
memset(r,0,sizeof r);
memset(l,0,sizeof l);
memset(u,0,sizeof u);
memset(d,0,sizeof d);
int s=0;

for(int i=0;i<f;i++)
{
int x,y;
cin>>x>>y;
m[x-1][y-1]=1;
s+=x-1+y-1;
}
int g[1001];
memset(g,0,sizeof g);
for(int i=0;i<b;i++)
for(int j=0;j<a;j++)
{
g+=m[j];
}
l[0]=0;
for(int i=1;i<b;i++)
{
l=l[i-1]+g[i-1];

}
for(int i=b-1;i>=0;i--)
r+=f-l;
/* for(int i=0;i<b;i++)
cout<<" righ "<<i<<" : "<<r;
cout<<endl;
for(int i=0;i<b;i++)
cout<<" left "<<i<<" : "<<l;
cout<<endl;*/
memset(g,0,sizeof g);
for(int i=0;i<a;i++)
for(int j=0;j<b;j++)
{
g+=m[j];
}
u[0]=0;
for(int i=1;i<a;i++)
u+=u[i-1]+g[i-1];
d[a-1]=g[a-1];
for(int i=a-2;i>=0;i--)
d[i]=f-u[i];
int mini=s;
int x=0,y=0;
int dis[1001];
dis[0]=s;
for(int i=1;i<a;i++)
{
dis[i]=dis[i-1]+u[i]-d[i];
}
int dd[1001];
for(int i=0;i<a;i++)
{
dd[0]=dis[i];
if(dis[i]<mini && m[i][0]==1)
{
mini=dis[i];
x=i;
y=0;
}
for(int j=1;j<b;j++)
{
dd[j]=dd[j-1]+l[j]-r[j];
if(dd[j]<mini && m[i][j]==1)
{
mini=dd[j];
x=i;
y=j;
}
}
}

cout<<"(Street: "<<x+1<<", Avenue: "<<y+1<<")"<<endl;

}

return 0;

}

Pleeeeeeeeeese HELP ME
Is there anyone solved this problem with sorting methode(and writhing the middle element of the sorted array)???[/cpp]

[little joey]

Is there anyone solved this problem with sorting methode(and writhing the middle element of the sorted array)???

Yes, that's the way to do it. Basically it's a two dimensional 'Vito's Family' (I just remember the title of the problem, not the number). In the discussion of that problem on the board, there is an excellent explanation (by Whinii F., if I'm correct) why the median method works in these cases.

[go]

At last I Accepted this problem, I understood how to write it with sorting.

[WR]

So what's wrong with howardchengs's algorithm?

I don't talk about array size, integer type and so on - just the algorithm.

added Nov 22

Perhaps somebody can tell me what's wrong with my program?

Code: Select all

#include <stdio.h>            /* scanf, printf */
#include <stdlib.h>           /* qsort */

/*******************************
 * ACM Contest - Problem 855   *
 * WA           *
 *******************************/

long STREETS[50002];
long AVENUES[50002];

int cmp(const void *ap, const void *bp)
{
  long a = *(long *)ap;
  long b = *(long *)bp;
  if (a < b) return -1;
  if (a > b) return 1;
  return 0;
}

int main(void)
{
  int nc, cases;
  int ns, streets;
  int na, avenues;
  int nf, friends;
  long medst;
  long medav;
  scanf("%d",&cases);
  for (nc=0;nc<cases;nc++){
    scanf("%d %d %d",&streets,&avenues,&friends);
    for (nf=0;nf<friends;nf++)
      scanf("%ld %ld",&STREETS[nf],&AVENUES[nf]);
    if (friends == 1){
      medst = STREETS[0];
      medav = AVENUES[0];
    }  
    else if (friends == 2){
      medst = (STREETS[0]+STREETS[1])/2;
      medav = (AVENUES[0]+AVENUES[1])/2;
    }
    else{
      qsort(STREETS,friends,sizeof(long),cmp);
      qsort(AVENUES,friends,sizeof(long),cmp);
      if ((friends % 2) == 0){
        medst = (STREETS[friends/2]+STREETS[friends/2-1])/2;
        medav = (AVENUES[friends/2]+AVENUES[friends/2-1])/2;
      }
      else{
        medst = STREETS[friends/2];
        medav = AVENUES[friends/2];
      }
    }
    printf("(Street: %ld, Avenue: %ld)\n",medst,medav);
  }
  return 0;
}

[Yaron Wittenstein]

Code: Select all

#include <stdio.h>

#define   MAXN   1000

int median(int a[], int n) /* a[i] <= MAX_VAL */
{
  int i, sum, m;

  sum = 0;
  i = 1;
  m = (n % 2 == 0 ? (n / 2 - 1) : n / 2);

  while (sum < m) {
    sum += a[i];
    i++;
  }

  return i;
}

void init(int a[], int n)
{
  int i;
  for(i = 1; i <= n; i++)
    a[i] = 0;
}

int main()
{
  int y[MAXN + 1];
  int x[MAXN + 1];
  int T, A, S, F;
  int median_x, median_y;
  int i;
  int street, avenue;

  scanf("%d", &T);
  for(; T > 0; T--) {
    scanf("%d %d %d", &S, &A, &F);
    init(x, A);
    init(y, S);

    for(i = 0; i < F; i++) {
      scanf("%d %d", &street, &avenue);
      y[street]++;
      x[avenue]++;
    }

    median_x = median(x, F);
    median_y = median(y, F);
    printf("(Street: %d, Avenue: %d)\n", median_y, median_x);
  }

  return 0;
}

please try to help me out here!

[Larry]

How does your median work?

You can solve this problem by sorting too, it's fast enough.

[Raj Ariyan]

Hi Yaron,

I think your median function does not always give right answer for all cases. You can sort street and avenue and then just print out the median (Simple Median Knowledge n/2 or n/2-1). Bye.

[Jan]

I got Accepted after several wrong answers. And finally I found the bug
in my code.
My algorithm is simple....

Code: Select all

    1. First sort the streets.
    2. Sort the avenues.
    3. Suppose they are in an array p[][2],
        p[][0] is the street number, p[][1] is the avenue number
        starting from 1 to n (n is the total number)
    4. If n is odd assign j <- (n+1)/2
    5. Otherwise assign j <- n/2
    6. Print p[j][0] and p[j][1]......

[Donotalo]

i think i'm doing the same thing but still getting wrong answer. here is my sample input and output:

Code: Select all

22 22 6
1 5
1 4
1 3
2 1
3 10
3 9

Code: Select all

(Street: 1, Avenue: 5)

Code: Select all

10 10 1
5 5

Code: Select all

(Street: 5, Avenue: 5)

Code: Select all

2 2 2
1 1
2 2

Code: Select all

(Street: 1, Avenue: 1)

Code: Select all

7 7 11
1 2
1 7
2 2
2 3
2 5
3 4
4 2
4 5
4 6
5 3
6 5

Code: Select all

(Street: 3, Avenue: 4)

is there anything tricky? please help me!

[Jan]

The problem is in multiple input format. So, your sets will be like

Input:

Code: Select all

4
22 22 6 
1 5 
1 4 
1 3 
2 1 
3 10 
3 9
10 10 1 
5 5 
2 2 2 
1 1 
2 2
7 7 11 
1 2 
1 7 
2 2 
2 3 
2 5 
3 4 
4 2 
4 5 
4 6 
5 3 
6 5

Output:

Code: Select all

(Street: 1, Avenue: 4)
(Street: 5, Avenue: 5)
(Street: 1, Avenue: 1)
(Street: 3, Avenue: 4)

And your first case is also wrong.

[Donotalo]

i've arranged the multiple input set as u suggested. i used the previous style just to be crystal clear.

now, about the first case: my program sorts it like the following:

Code: Select all

1 3
1 4
1 5
2 1
3 9
3 10

why the median shud be (1 4) here?

[Jan]

If all the friends meet at (1, 4)

Code: Select all

1 3  => 0 + 1 = 1
1 4  => 0 + 0 = 0
1 5  => 0 + 1 = 1
2 1  => 1 + 3 = 4
3 9  => 2 + 5 = 7
3 10 => 2 + 6 = 8
Total Distance => 21

If all the friends meet at (1, 5)

Code: Select all

1 3  => 0 + 2 = 2
1 4  => 0 + 1 = 1
1 5  => 0 + 0 = 0
2 1  => 1 + 4 = 5
3 9  => 2 + 4 = 6
3 10 => 2 + 5 = 7
Total Distance => 21

So, they both are optimal ones. But the problem states

If there is more than one candidate point, the rule imposes that the meeting point is the one corresponding to the smaller number for street and avenue.

(1, 4) is smaller than (1, 5). So, the output should be (1, 4).
And if you are following my algorithm (as I described before) then I think that you haven't understood it correctly.

[Donotalo]

yeah, u r right. i don't understand ur algorithm and cudn't find any mistake in my algorithm to me, both of the algorithms are same. i've corrected the flaw and still getting wrong answer.

can anyone give me any tricky input?

[Jan]

After sorting you will get

Code: Select all

1 1
1 3
1 4
2 5
3 9
3 10

from the above set. . And the median is (1, 4). Hope you understand my algorithm.