/**
* Exports a distance matrix to a text file
* @param d Pointer to matrix
* @param fa Feature vector array
* @param file File name
*/
void export_dist(double *d, farray_t *fa, const char *file)
{
assert(d && fa && file);
int i, j;
FILE *f;
if (verbose > 0)
printf("Exporting distance matrix to '%s'.\n", file);
if (!(f = fopen(file, "w"))) {
error("Could not create file '%s'.", file);
return;
}
/* Print version header */
malheur_version(f);
/* Print distance header */
fprintf(f, "# ---\n# Distance matrix for %s\n", fa->src);
fprintf(f, "# Matrix size: %lu x %lu\n# ---\n", fa->len, fa->len);
fprintf(f, "# <report> <cluster> <dist1> <dist2> ... <distn>\n");
/* Print matrix */
for (i = 0; i < fa->len; i++) {
fprintf(f, "%s %s ", fa->x[i]->src, farray_get_label(fa, i));
for (j = 0; j < fa->len; j++)
fprintf(f, "%g ", d[i * fa->len + j]);
fprintf(f, "\n");
}
fclose(f);
}
/**
* Return an array of rejected feature vectors
* @param c Cluster structure
* @param f Array of feature vectors
* @return Rejected feature vectors
*/
farray_t *cluster_get_rejected(cluster_t *c, farray_t *f)
{
int i;
farray_t *r = farray_create("rejected");
for (i = 0; i < f->len; i++) {
if (c->cluster[i])
continue;
farray_add(r, fvec_clone(f->x[i]), farray_get_label(f, i));
}
return r;
}
/**
* Exports classification results
* @param p Prototype structure
* @param fa Feature vector array
* @param as Assignments to protoypes
* @param file File name
*/
void export_class(farray_t *p, farray_t *fa, assign_t *as, const char *file)
{
assert(p && fa && file);
int i, j;
char *l;
FILE *f;
if (verbose > 0)
printf("Exporting classification to '%s'.\n", file);
if (!(f = fopen(file, "w"))) {
error("Could not create file '%s'.", file);
return;
}
/* Print version header */
malheur_version(f);
/* Evaluate some quality functions */
double *e = quality(fa->y, as->label, as->len);
/* Print prototype header */
fprintf(f, "# ---\n# Classification for %s\n", fa->src);
fprintf(f, "# Precision of classification: %4.1f %%\n",
e[Q_PRECISION] * 100.0);
fprintf(f, "# Recall of classification: %4.1f %%\n",
e[Q_RECALL] * 100.0);
fprintf(f, "# F-measure of classification: %4.1f %%\n",
e[Q_FMEASURE] * 100.0);
fprintf(f, "# ---\n# <report> <label> <prototype> <distance>\n");
for (i = 0; i < fa->len; i++) {
j = as->proto[i];
l = as->label[i] ? farray_get_label(p, j) : "rejected";
fprintf(f, "%s %s %s %g\n", fa->x[i]->src, l, p->x[j]->src,
as->dist[i]);
}
fclose(f);
}
/**
* Simple test cases classification
*/
int test_classify()
{
int i, k, err = 0;
fvec_t *f;
test_printf("Classification using prototypes");
/* Prepare training data */
farray_t *fa1 = farray_create("train");
for (i = 0; train_data[i].str; i++) {
f = fvec_extract(train_data[i].str, strlen(train_data[i].str), NULL);
farray_add(fa1, f, train_data[i].label);
}
/* Prepare testing data */
farray_t *fa2 = farray_create("train");
for (i = 0; test_data[i].str; i++) {
f = fvec_extract(test_data[i].str, strlen(test_data[i].str), NULL);
farray_add(fa2, f, test_data[i].label);
}
/* Classification of test data */
config_set_float(&cfg, "classify.max_dist", 1.41);
assign_t *a = class_assign(fa2, fa1);
/* Check predicted labels */
for (k = 0; test_data[k].str; k++) {
char *l = farray_get_label(fa1, a->proto[k]);
err += strcmp(l, test_data[k].label) != 0;
}
/* Clean up */
assign_destroy(a);
farray_destroy(fa1);
farray_destroy(fa2);
test_return(err, i);
return err;
}
/**
* Exports results from the incremental analysis (phase 1). The results are
* obtained by first classifying and then clustering reports. In the first
* phase the classified reports are written to the output file.
* @param p Prototype struture
* @param fa Feature vector array
* @param as Assignments to prototypes
* @param file File name
*/
void export_increment1(farray_t *p, farray_t *fa, assign_t *as,
const char *file)
{
int i, j;
FILE *f;
if (verbose > 0)
printf("Exporting incremental analysis (1) to '%s'.\n", file);
if (!(f = fopen(file, "w"))) {
error("Could not create file '%s'.", file);
return;
}
/* Print version header */
malheur_version(f);
/* Print incremental header */
fprintf(f, "# ---\n# Incremental analysis for %s\n", fa->src);
fprintf(f, "# ---\n# <report> <cluster> <prototype> <distance>\n");
if (!p || !as) {
fclose(f);
return;
}
for (i = 0; i < fa->len; i++) {
if (!as->label[i])
continue;
j = as->proto[i];
fprintf(f, "%s %s %s %g\n", fa->x[i]->src, farray_get_label(p, j),
p->x[j]->src, as->dist[i]);
}
fclose(f);
}
