Tag: largest

Some guess

The code you provide is missing important information that we must guess.

Being a programmer means being able to present the problem and the solution without ambiguity. Here is one important and some pedantic ambiguities listed.

• What type are the items referenced by array?

  We can assume they are int but depending on the compiler settings your code could still work for other types. The problem may be hidden behind warnings that we can not see.

  If for example the array contains unsigned int then adding what is a positive value (in the array) can be subtracted if the top bit is 1

• What is row_count, is it a static typed value, a defined value (has no type), or …?

• You provided no library references, eg #include <stdio.h> Yes it is a very standard include associated with printf but we are still left to guess this is a fact.

Rewrite

Taking your code at face value

Two bugs

1. Sum will sum all rows as you do not reset its value for each row.

2. The max value temp is initialized with the wrong value. It should have the lowest possible value. That means that largest_row may never get a value.

   Depending on the compiler setting what is printed could be a random value, or zero.

Poor design

The code is inefficient and the naming is poor.

• largest_row is a function so give it a capital Largest_row

• largest_row is not what the function does. It prints the index of the largest row. Using PrintIdxLargestRowSum would be better however it is too long. If we make the function return the largest row, the name could be IdxLargestSum

• It is a little odd as you are checking for a max value as you are summing the row.

  First sum the row, then check for the max value.

• Keep variables scoped as close as possible to the code they are used in. Declare sum inside the first loop.

Using your code as a template we get

// Assumes global row_count > 0
int Idx_largest_row(void) {
    int idx, max;
    for (int i = 0; i < row_count; i++) {
        int sum = 0
        for (int j = 0; j < row_count; j++) {
            sum += array(i)(j); 
        }
        if (i == 0 || (i > 0 && sum > max)) {
            max = sum;
            idx = i;
        }
    }
    return idx;
}

// use as
printf("%in", Idx_largest_row());

Example

The above rewrite is still rather poor in my view.

• There should be two functions, one to sum a row the other to find the max sum.

• The array and row counts should be an argument

• The inner arrays should be of defined length

#include <stdio.h>

#define COL_COUNT  5
#define uint unsigned int

int rows()(COL_COUNT) = {
    {10,   2, 320, 344, -130},
    {30,  12, 330, 444, -160},
    {100,  2, 123, 207,   50},
    {102, 82, 530, 544, -150},
    {103, 72, 730, 444, -140}
};

int SumRow(int *row, uint count) {
    int sum = 0;
    for (uint i = 0; i < count; i++) { sum += row(i); }
    return sum;
}
uint LargestRowIdx(int rows()(COL_COUNT), uint rowCount) {
    uint idx = 0;
    int max = SumRow(rows(0), COL_COUNT);
    for (uint i = 1; i < rowCount; i++) {
        int sum = SumRow(rows(i), COL_COUNT);
        if (sum > max) { 
            max = sum; 
            idx = i;
        }
    }
    return idx;
}

void main() {
    printf("Index of largest sum: %un", LargestRowIdx(rows, 5));
    return;
}

_{Note that the number of rows is a magic number 5, however you should use a defined value or variable}

_{Note I use unsigned int when indexing and rather than have to type all that I use the the defined #define uint unsigned int. There is no reason you can not use an int for indexes}

_{Note I use camelCase names rather than snake_case. Which to use is up to you}