Thread: How do I fix this piece of code?

The output I get for each year range is inf. I'm trying to read into the file and calculating the frequency per hundred thousand, for a name in a specific year range. How can I fix this? Output:

Name?
julie
1921 - 1925: inf
1926 - 1930: inf
1931 - 1935: inf
1936 - 1940: inf
1941 - 1945: inf
1946 - 1950: inf
1951 - 1955: inf
1956 - 1960: inf
1961 - 1965: inf
1966 - 1970: inf
1971 - 1975: inf
1976 - 1980: inf
1981 - 1985: inf
1986 - 1990: inf
1991 - 1995: inf
1996 - 2000: inf
2001 - 2005: inf

Code:

Code:
#include <stdio.h>
#include <stdlib.h>
#include <string.h>


struct NameRecord {
        char name[31];
        int year;
        int frequency;
};


int getRawData(FILE *fp, struct NameRecord records[], int currSize);


int main() {


struct  NameRecord records[150000];


FILE    *fp = NULL;


int     size = 0, currSize = 0, yearRangeTotal[17], nameTotal[17], year = 0;
char    name[31];
double  perHundredThousand[17], frequency = 0;


        fp = fopen("femalebabynames.csv", "r");
        size = getRawData(fp, records, currSize);


        printf("Name?\n");
        scanf("%s", name);


        int i, x;


                memset(yearRangeTotal, 0, sizeof(yearRangeTotal));
                for(i = 0; i < size; i++) {
                        x = ((records[i].year - 1) / 5 - 384);
                        yearRangeTotal[x] =+ yearRangeTotal[x] + records[i].frequency;
                }


                memset(nameTotal, 0, sizeof(nameTotal));
                for(i = 0; i < size; i++) {
                        x = ((records[i].year - 1) / 5 - 384);
                        if(strncmp(name, records[i].name, 30) == 0) {
                                nameTotal[x] =+ nameTotal[x] + records[i].frequency;
                        }
                }


                memset(perHundredThousand, 0, sizeof(perHundredThousand));
                for(i = 0; i < 17; i++) {
                        perHundredThousand[i] = 100000.00 / nameTotal[i] * yearRangeTotal[i];
                }


                for(i = 0; i < 17; i++) {
                        printf("%.4d - %.4d: %.2f\n", (i+384)*5+1, (i+385)*5, perHundredThousand[i]);
                }


                printf("\n");


        return 0;


}


int getRawData(FILE* fp, struct NameRecord records[], int currSize) {
        while(EOF != fscanf(fp,"%d,%[^,],%d", &records[currSize].year, records[currSize].name, &records[currSize].frequency)) {
                currSize++;
        }
        return currSize;


}

I tried that but now the output is:
1921 - 1925: inf
1926 - 1930: -inf
1931 - 1935: -inf
1936 - 1940: -inf
1941 - 1945: inf
1946 - 1950: -inf
1951 - 1955: -inf
1956 - 1960: inf
1961 - 1965: -inf
1966 - 1970: inf
1971 - 1975: inf
1976 - 1980: inf
1981 - 1985: -inf
1986 - 1990: inf
1991 - 1995: inf
1996 - 2000: -inf
2001 - 2005: -inf

The smallest possible sample input file that demonstrates this problem would be immensely helpful.

Some notes in the mean time:

1. Don't use magic numbers. #define some constants. Why 31, 17, 384, 5, 385? Consider defining some constants in terms of others

Code:

#define YEAR_GROUP_SIZE 5
#define MAX_YEARS 85
#define NUM_YEAR_GROUPS (MAX_YEARS / YEAR_GROUP_SIZE + 1)  // +1 in case they don't divide evenly, and you need to deal with a partial year group

2. Does records really need to hold 150k entries?

3. Check the value of your file functions like fopen. If you can't open the file, print an error (look into the perror() function) and quit. You can't do anything else useful.

4. =+ and += are not the same thing

Code:

x =+ y;  // x = +y;  x is assigned the value of y
x += y;  // x = x + y;  x is assigned the value of x+y

5. Read the documentation for fscanf. You should check the return value to make sure you scanned all expected values. If not, you have a corrupted line in your file and all further parsing is suspect.

6. inf can come from a few things: divide by zero (or almost 0), random, uninitialized data, etc. Make sure you initialize everything properly.

7. On that note, your memset() calls will set the memory to all-bits zero. That may not correspond to an actual value of 0 in a floating point variable (like your array of doubles 'perHundredThousand'). Explicitly set each element to 0.0 with a loop. Since you do this several times, writing a function would be a good idea.

8. There is no point in passing in currSize to the getRawData function. Just make it a local, initialized to 0, and return it when done.

That should give you something to work on for a while.

Originally Posted by oogabooga

On a second look, changing to += is still wrong.
You have

Code:

        yearRangeTotal[x] += yearRangeTotal[x] + records[i].frequency;

You want

Code:

        yearRangeTotal[x] += records[i].frequency;

Ditto for nameTotal.

Haha! On third look, if that was the case, then the original should have worked since it was adding the current value of yearRangeTotal with records.frequency and storing it back into yearRangeTotal.

Here's a sample of what's in the data:
Beginning

2005	aaliya	7
2006	aaliya	5
2007	aaliya	5
2008	aaliya	7
2009	aaliya	5

End

2010	ZUNAIRA	5
2008	ZUNAIRAH	6
2010	ZUNAIRAH	7

Give me a moment while I try to make sense of everything... sorry I'm still fairly new at programming. I have a sense of what I'm trying to do but I don't think I'm using the right functions and format

Originally Posted by oogabooga

I see nothing wrong with the way he's using strncmp.

Yeah, I must have been editing my comment when you were posting this

The file does read with my code because I tested it with:

Code:

int getRawData(FILE* fp, struct NameRecord records[], int currSize) {
        while(EOF != fscanf(fp,"%d, %[^,], %d", &records[currSize].year, records[currSize].name, &records[currSize].frequency)) {
               printf("%d  ---  %d --- %s --- %d\n", currSize, records[currSize].year, records[currSize].name, records[currSize].frequency);
               currSize++;
        }
        return currSize;


}

Originally Posted by deepee

The file does read with my code because I tested it with:

Code:

int getRawData(FILE* fp, struct NameRecord records[], int currSize) {
        while(EOF != fscanf(fp,"%d, %[^,], %d", &records[currSize].year, records[currSize].name, &records[currSize].frequency)) {
               printf("%d  ---  %d --- %s --- %d\n", currSize, records[currSize].year, records[currSize].name, records[currSize].frequency);
               currSize++;
        }
        return currSize;


}

If that's the case, then the input file you posted does not match the input file on your system. fscanf will try to match those commas, and fail if it can't. Also, it's worth pointing out that the code above does not match the code in your first post, since you have spaces after the commas in your format string. We need to know exactly what you're working with, if we are to help you.