1. #include<stdio.h>
#include<math.h>
#include<string.h>
void copy(char a[50],char b[100][10][50],int u,int w);
float toi(char M[100][10][50],int i,int k);
int stcmp(char M[100][10][50],int i,int k,char (*str)[20],int f);
int missing_handle(float d[100][50],float new_d[100][50],int w,int u,int missing_attribute[],int
missing_tuple[],int datatype[]);
float detrm(float[25][25],float [25][25],int);
void cofact(float[25][25],float [25][25],int);
void trans(float[25][25],float[25][25],float [25][25],int);
void inverse(float [25][25],float ver[25][25],int);
void multiply(float [25][25],float [25],float [25][25],int);
void normalization(float d[100][50],int w,int u,int datatype[50]);
float mode(float new_d[100][50],int w,int v);
void missing_check(float d[100][50],int w,int u,int datatype);
int main()
{
FILE *fp;
int m,i=0,j=0,k=0,u=0,v=0,w=0,loc=0,f=0,norm;
char ch,name[50],a[100],M[100][10][50],*NomiValue[50];
int Nattributes,datatype[50],Nomnmber;
char (*str)[20],r=0;
float d[100][50],value;
printf("Enter the total number of attrributes=");
scanf("%d",&Nattributes);
for(i=0;i<Nattributes;i++)
{
printf("Enter the datatype of %dth attribute",i+1);
printf("ntNumerial=0 Nominal=1t=");
scanf("%d",&datatype[i]);
if(datatype[i]==0)
{
continue;
}
else
{
printf("nntEnter the total number of possible values of %dth attribute=",i+1);
scanf("%d",&Nomnmber);
printf("Enter all the possible attribute values ");
str=(char (*)[20])malloc(15*sizeof(*str));
NomiValue[r]=str;

2. r++;
for(j=0;j<Nomnmber;j++)
{
printf("n%d Attribute value=",j+1);
scanf("%s",str[j]);
}
}
}
/*for(k=0;k<2;k++)
{
str=NomiValue[k];
for(i=0;i<2;i++)
{
printf("%st",str[i]);
}
printf("n");
}*/
i=0;
//SEARCHING FOR TEXT FILE
do
{
puts("ntEnter the path of the file(EXTENSION SHOULD BE .txt): ");
gets(name);
//OPENING FILE IN READ MODE
fp=fopen(name,"r");
if(fp=='0')
puts("ntFile not found... Re-enter:");
else
puts("ntFile found");
}while(fp=='0');
//READING FROM FILE CHARACTER BY CHARACTER
while(1)
{
//WHILE BEGINS
ch=fgetc(fp);
if(ch==EOF)
{
printf("world");
break;
}
if(ch==32&&i==0)

3. {
//printf("i m reachable");
M[u][v][w]=EOF;
v=0;
u++;
ch=fgetc(fp);
}
else if(ch==32||ch=='n')
{
a[i]='0';
copy(a,M,u,w);
if(ch=='n')
{
i=0;
u=0;
v=0;
w++;
}
else{
i=0;
u++;
v=0;
continue;
}
}
else{
a[i]=ch;
i++;
}
}//end of while
printf("ntotal no. of tuples=%d",w+1);
printf("ntotal no. of attributes=%dn",u);
//printf("hellon");
for(k=0;k<=w;k++)
{
for(i=0;i<u;i++)
{
j=0;
while(M[i][j][k]!='0')
{
printf("%c",M[i][j][k]);

4. j++;
}
printf("t");
}
printf("n");
}
//printf("*******%c",M[0][0][0]);
//Conversion of 3D Array to integer array 2D
for(k=0;k<=w;k++)
{ j=0;
r=0;
loc=0;
printf("n");
for(i=0;i<Nattributes;i++)
{
j=0;
loc=0;
if(datatype[i]==0)
{
value=toi(M,i,k);
d[k][i]=value;
//printf("%f",value);
}
else
{
str=NomiValue[r];
r++;
for(f=0;f<15;f++)
{
// printf("n########%s###### %d#######n",str[f],stcmp(M,i,k,str,f));
if(stcmp(M,i,k,str,f)==0)
{
//printf("t%d",f);
d[k][i]=f;
loc=1;
j++;
break;
}
}
if(loc==0)
{

5. d[k][i]=-1;
}
}
}
}
fclose(fp);
//printf("nnvatesh");
for(i=0;i<=w;i++)
{
for(j=0;j<Nattributes;j++)
printf("%ft",d[i][j]);
printf("n");
}
missing_check(d,w,u,datatype);
printf("Do you want to normalise the data");
if(scanf("%d",&norm))
{
normalization(d,w,u,datatype);
}
return 0;
}
float toi(char M[100][10][50],int i,int k)
{
int j=0,e=0;
int count=0,countf=0;
float sum=0;
int setNum=0,c;
while(M[i][j][k]!='0')
{
if(M[i][j][k]=='.')
{
setNum=1;
break; //means numeric is float
}
j++;
}
j=0;
if(setNum==0)
{
while(M[i][j][k]!='0')
{
count++;

6. j++;
}
}
else
{
while(M[i][j][k]!='.')
{
countf++;
j++;
}
}
//printf("count=%d",count);
j=0;
while(M[i][j][k]!='0')
{
e=0;
if(M[i][j][k]==EOF)
return -1;
if(setNum==1)
{
if(M[i][j][k]!='.'){
c=M[i][j][k]-48;
sum=sum+c*(pow(10,countf-1-j));
j++;
}
else{
e++;
j++;
while(M[i][j][k]!='0')
{
c=M[i][j][k]-48;
sum=sum+c/(float)(pow(10,e));
e++;
j++;
}
break;
}
}
else{
if(M[i][j][k]!='.'){

7. c=M[i][j][k]-48;
sum=sum+c*(pow(10,count-1-j));
j++;
}
}
}
//printf("n%d",sum);
return sum;
}
void copy(char a[50],char b[100][10][50],int u,int w)
{
//printf("copy");
int j=0;
while(a[j]!='0')
{
// printf("n%c",a[j]);
b[u][j][w]=a[j];
j++;
}
b[u][j][w]='0';
}
int stcmp(char M[100][10][50],int i,int k,char (*str)[20],int f)
{
if(M[i][0][k]==str[f][0])
{
return 0;
}
else
return 1;
}
void missing_check(float d[100][50],int w,int u,int datatype)
{
int missing_attribute[50],missing_tuple[100],i,j,handle,loc=0;
float new_d[100][50];
int m=0,n=0;
//finding missing attributes
printf("nNumber of attributes=%dn",u);
printf("Number of tuples=%dn",w+1); //note total no. of tuples in data is w+1
for(j=0;j<u;j++)
{
loc=0;

8. for(i=0;i<=w;i++)
{
if(d[i][j]==-1)
{
loc=1;
break;
}
}
if(loc==1)
{
missing_attribute[j]=1;
//printf("****%d",missing_attribute[j]);
}
else{
missing_attribute[j]=0;
}
}
//finding missing tuples
for(i=0;i<=w;i++)
{
loc=0;
for(j=0;j<u;j++)
{
if(d[i][j]==-1)
{
loc=1;
break;
}
}
if(loc==1){
missing_tuple[i]=1;
}
else
{
missing_tuple[i]=0;
}
}
loc=0;
for(i=0;i<u;i++)
{
if(missing_attribute[i]==1)

9. {
printf("%d attribute contains missing datann ",i+1);
loc=1; //it is correct there is no need to change the value of loc=0
}
}
if(loc==1)
{
printf("Want to ignore missing values or handle using imputation methodn");
printf("Enter 0 for ignoring and 1 for handling=");
scanf("%d",&handle);
}
for(i=0;i<=w;i++)
{
n=0;
if(missing_tuple[i]!=1)
{
for(j=0;j<u;j++)
{
new_d[m][n]=d[i][j];
n++;
}
m++;
}
}
//printf("*******m=%d*********n=%d*********",m,n);
if(handle==1)
{
missing_handle(d,new_d,w,u,missing_attribute,missing_tuple,datatype);
}
else
{
for(i=0;i<m;i++)
{
for(j=0;j<n;j++)
{
printf("%ft",new_d[i][j]);
}
printf("n");
}
}
}

10. int missing_handle(float d[100][50],float new_d[100][50],int w,int u,int missing_attribute[],int
missing_tuple[],int datatype[]) //yth attribute contain missing data
{
int i=0,j=0,add=0,m=0,n=0,k=0,c=0,loc=0,v,f,g,z=0;
float sum,N[100][50],A[100][50],variable; //program can handle maximum 50 attributes
float (*p)[50];
int dependent[50],independent[50];
int count=0,r,s,q;
float (*array1)[25],*array2,ver[25][25],mul[25][25];
//calculation of no. of tuples after ignoring missing tuples
for(i=0;i<=w;i++)
{
if(missing_tuple[i]!=1)
count++;
}
printf("nno. of tuples after ignoring missing tuples=%dn",count);
for(i=0;i<u;i++)
{
if(missing_attribute[i]==1){
dependent[m]=i;
m++;
}
else{
independent[n]=i;
n++;
}
}
/* for(i=0;i<n;i++)
printf("%dt",independent[i]);
printf("n");*/
/* for(i=0;i<n;i++)
printf("%dt",independent[i]);
printf("n");*/
//formation of matrix with all independent variables and only one dependent variable
printf("nm=%d n=%d",m,n);
//for(z=0;z<m;z++) //will work depend on number of dependent variable
// {
v=dependent[z];
printf("nv=%d",v);
p=(float (*)[50])malloc(100*sizeof(*p));
printf("vatesh developer");

11. for(k=0;k<count;k++)
{
c=0;
for(j=0;j<n;j++)
{
f=independent[j];
//printf("#####%d",f);
p[k][c]=new_d[k][f];
//printf("%ft",p[k][c]);
c++;
}
p[k][c]=new_d[k][v];
//printf("%ft",p[k][c]);
c++;
//printf("n");
}
printf("ntuples after removaln");
for(r=0;r<count;r++) //no. of tuple without any missing value
{
for(s=0;s<=n;s++)
{
printf("%ft",p[r][s]); //p[][] contains independent variable and one dependent variable
}
printf("n");
}
//concept to be checked 2mrw
for(g=0;g<=n;g++)
{
for(s=g;s<=n;s++)
{
sum=0;
for(i=0;i<count;i++)
{
sum=sum+(p[i][g])*(p[i][s]);
}
printf("%f ",sum);
A[g+1][s+1]=sum;
}
}
N[0][0]=count;
for(j=0;j<=n;j++)

12. {
sum=0;
for(k=0;k<count;k++)
{
sum=sum+p[k][j];
}
N[0][j+1]=sum;
}
for(i=1;i<=n+1;i++)
{
for(j=i;j<=n+1;j++)
{
N[i][j]=A[i][j];
}
}
printf("n");
/*for(i=0;i<=n+1;i++)
{
for(j=0;j<=n+1;j++)
{
if(i<=j)
//printf("%ft",N[i][j]);
else
//printf("0.000000t");
}
printf("n");
}*/
//taking transpose of matrix of N
for(i=0;i<=n+1;i++)
{
for(j=i;j<=n+1;j++)
{
N[j][i]=N[i][j];
}
printf("n");
}
for(i=0;i<=n+1;i++)
{
for(j=0;j<=n+1;j++)

13. {
printf("%ft",N[i][j]);
}
printf("n");
}
//Array is the final array1,take its inverse and multiply by array2 to get final coefficients of equations.
array1=(float (*)[25])malloc(sizeof(*array1)*n);
array2=(float *)malloc(sizeof(float)*n);
for(i=0;i<=n;i++)
{
for(j=0;j<=n;j++)
{
array1[i][j]=N[i][j];
}
}
printf("nArray1n");
for(i=0;i<=n;i++)
{
for(j=0;j<=n;j++)
{
printf("%ft",array1[i][j]);
}
printf("n");
}
printf("nArray2n");
for(i=n+1;i<=n+1;i++) //will work only once
{
for(j=0;j<n+1;j++)
{
array2[j]=N[i][j];
printf("%ftn",array2[j]);
}
printf("n");
}
inverse(array1,ver,n+1);
printf("**********************************n");
for(i=0;i<=n;i++)
{
for(j=0;j<=n;j++)
{

14. printf("%ft",ver[i][j]);
}
printf("n");
}
multiply(ver,array2,mul,n+1); //note:-order of matrix is n+1
for(i=0;i<n+1;i++)
{
for(j=0;j<1;j++) //will work only once hence can be eliminated
{
printf("%ft",mul[i][j]);
}
printf("n");
}
for(i=0;i<w+1;i++)
{
sum=0;
if(d[i][v]==-1)
{
for(j=0;j<n;j++)
{
q=independent[j];
//printf("q=%")
sum=sum+d[i][q]*mul[j+1][0];
}
sum=sum+mul[0][0];
if(datatype[v]!=1)
{
d[i][v]=sum;
printf("nd[%d][%d]=%f",i,v,sum);
}
else
{
variable=mode(new_d,w,v);
d[i][v]=variable;
printf("nd[%d][%d]=%f",i,v,d[i][v]);
}
}

15. }
for(i=0;i<25;i++)
{
free((void *)array1[i]);
}
free(array1);
free(array2);
for(i=0;i<100;i++)
{free(p[i]);}
free(p);
missing_check(d,w,u,datatype);
//}
}
/* Inverse of a n by n matrix */
void inverse(float a[25][25],float inv[25][25],int k)
{
float d;
int i,j;
/*printf("in inversen");
for(i=0;i<k;i++)
{
for(j=0;j<k;j++)
{
printf("%ft",a[i][j]);
}
printf("n");
}
printf("k=%d",k);*/
d=detrm(a,inv,k);
printf("THE DETERMINANT IS=%f",d);
if(d==0)
printf("nMATRIX IS NOT INVERSIBLEn");
else

16. cofact(a,inv,k);
}
/******************FUNCTION TO FIND THE DETERMINANT OF THE
MATRIX************************/
float detrm(float a[25][25],float inv[25][25],int k)
{
float s=1,det=0,b[25][25];
int i,j,m,n,c;
/*
printf("nnIn determinantn");
for(i=0;i<k;i++)
{
for(j=0;j<k;j++)
{
printf("%ft",a[i][j]);
}
printf("n");
}*/
if(k==1)
{
return(a[0][0]);
}
else
{
det=0;
for(c=0;c<k;c++)
{
m=0;
n=0;
for(i=0;i<k;i++)
{
for(j=0;j<k;j++)
{
b[i][j]=0;
if(i!=0&&j!=c)
{
b[m][n]=a[i][j];
if(n<(k-2))
n++;
else

17. {
n=0;
m++;
}
}
}
}
det=det+s*(a[0][c]*detrm(b,inv,k-1));
s=-1*s;
}
}
return(det);
}
/*******************FUNCTION TO FIND COFACTOR*********************************/
void cofact(float num[25][25],float inv[25][25],int f)
{
float b[25][25],fac[25][25];
int p,q,m,n,i,j;
for(q=0;q<f;q++)
{
for(p=0;p<f;p++)
{
m=0;
n=0;
for(i=0;i<f;i++)
{
for(j=0;j<f;j++)
{
b[i][j]=0;
if(i!=q&&j!=p)
{
b[m][n]=num[i][j];
if(n<(f-2))
n++;
else
{
n=0;

18. m++;
}
}
}
}
fac[q][p]=pow(-1,q+p)*detrm(b,inv,f-1);
}
}
trans(num,fac,inv,f);
}
/*************FUNCTION TO FIND TRANSPOSE AND INVERSE OF A
MATRIX**************************/
void trans(float num[25][25],float fac[25][25],float inv[25][25],int r)
{
int i,j;
float b[25][25],d;
printf("In Inverse");
for(i=0;i<r;i++)
{
for(j=0;j<r;j++)
{
b[i][j]=fac[j][i];
}
}
d=detrm(num,inv,r);
printf("n**************j =%dn",j);
inv[i][j]=0;
for(i=0;i<r;i++)
{
for(j=0;j<r;j++)
{
inv[i][j]=b[i][j]/d;
}
}
printf("nTHE INVERSE OF THE MATRIX:n");

19. for(i=0;i<r;i++)
{
for(j=0;j<r;j++)
{
printf("t%f",inv[i][j]);
}
printf("n");
}
}
void multiply(float ver[25][25],float array2[25],float mul[25][25],int k)
{
int i,j,m=0;
float sum=0;
for(i=0;i<k;i++)
{
for(j=0;j<1;j++) //will work only once hence can be eliminated
{
for(m=0;m<k;m++)
{
sum=sum+ver[i][m]*array2[m];
}
mul[i][j]=sum;
}
}
}
normalization(float d[100][50],int w,int u,int datatype[50])
{
float min,maxA,minA;
int i,j,norm=0,loc=0;
for(j=0;j<u;j++)
{
norm=0;
// loc=0;
/* if(datatype[j]==0);
{
for(i=0;i<=w;i++)
{
if(d[i][j]<0)
{loc=1;
break;
}

20. }
}
if(loc==1)
{
min=d[0][j];
for(i=0;i<=w;i++)
{
if(d[i][j]<min)
{
min=d[i][j];
}
}
for(i=0;i<=w;i++)
{
d[i][j]=d[i][j]-min;
}
}
*/
for(i=0;i<=w;i++)
{
if(d[i][j]>1||d[i][j]<0)
{
norm=1;
break;
}
}
if(norm==1)
{
minA=d[0][j];
for(i=0;i<=w;i++)
{
if(d[i][j]<minA)
{
minA=d[i][j];
}
}
maxA=d[0][j];
for(i=0;i<=w;i++)
{
if(d[i][j]>maxA)
{
maxA=d[i][j];

21. }
}
for(i=0;i<=w;i++)
{
d[i][j]=(d[i][j]-minA)/(maxA-minA);
}
}
}
for(i=0;i<=w;i++)
{
for(j=0;j<u;j++)
printf("%ft",d[i][j]);
printf("n");
}
}
float mode(float new_d[100][50],int w,int v)
{
float my[100],occurence[100],max;
int e=0,i,j,number,ret;
for(i=0;i<=w;i++)
{
for(j=0;j<e;j++)
{
if(my[j]==new_d[i][v])
{
continue;
}
my[e]=new_d[i][v];
e++;
}
}
for(i=0;i<e;i++)
{
for(j=0;j<=w;j++)
{
if(my[i]==new_d[j][v])
{

22. number++;
}
}
occurence[i]=number;
}
//now calculate max occurence
max=occurence[0];
ret=0;
for(j=0;j<e;j++)
{
if(max<occurence[j])
{
max=occurence[j];
ret=j;
}
}
return new_d[ret][v];
}