read a matrix of floats in tsv format

read_float_tsv.c

typedef struct _FloatMatrix {
  uint32_t N;
  uint32_t M;
  float *data;
} FloatMatrix;


/* read matrix from tsv file */
/* matrix is allocated, will need to be freed */
FloatMatrix * read_txt_float_matrix(char *fpath)
{
  uint32_t R, C;
  R = 0; C = 0;

  FILE *fp = fopen(fpath, "r");
  if(!fp)
  {
    fprintf(stderr, "cannot open %s for reading\n", fpath);
    return NULL;
  }

  // number of columns
  rewind(fp);
  uint32_t first_line_len = 0;
  while(1)
  {
    char c = fgetc(fp);
    if(c == '\n')
    {
      C++;
      break;
    }
    else if(c == '\t')
      C++;
    first_line_len++;
  }
  rewind(fp);

  uint32_t n_max_rows = 4096;
  FloatMatrix *X = (FloatMatrix *) malloc(sizeof(FloatMatrix));
  X->data = (float *) malloc(sizeof(float) * n_max_rows * C);

  uint32_t i = 0;
  while(!feof(fp))
  {
    for(uint32_t j = 0; j < C; j++)
    {
      fscanf(fp, "%f ", X->data + i*C + j);
      /* printf("%f ", *(X->data + i*C + j)); */
    }
    /* printf("\n"); */
    i++;
    if(i == n_max_rows)
    {
      /* printf("reallocing\n"); */
      n_max_rows *= 2;
      X->data = realloc(X->data, sizeof(float) * n_max_rows * C);
    }
  }

  X->N = i;
  X->M = C;
  // TODO trim the data ... (not strictly necessary)
  /* printf("first line len is %d, matrix is: %d x %d\n", first_line_len, X->N, X->M); */
  fclose(fp);
  return X;
}