if (mi->B[i].h) n += kh_size((idxhash_t*)mi->B[i].h);
for (i = 0; i < 1U<<mi->b; ++i) {
idxhash_t *h = (idxhash_t*)mi->B[i].h;
khint_t k;
if (h == 0) continue;
for (k = 0; k < kh_end(h); ++k)
if (kh_exist(h, k)) {
sum += kh_key(h, k)&1? 1 : (uint32_t)kh_val(h, k);
if (kh_key(h, k)&1) ++n1;
}
}
fprintf(stderr, "[M::%s::%.3f*%.2f] distinct minimizers: %d (%.2f%% are singletons); average occurrences: %.3lf; average spacing: %.3lf; total length: %ld\n",
__func__, realtime() - mm_realtime0, cputime() / (realtime() - mm_realtime0), n, 100.0*n1/n, (double)sum / n, (double)len / sum, (long)len);
}
int mm_idx_index_name(mm_idx_t *mi)
{
khash_t(str) *h;
uint32_t i;
int has_dup = 0, absent;
if (mi->h) return 0;
h = kh_init(str);
for (i = 0; i < mi->n_seq; ++i) {
khint_t k;
k = kh_put(str, h, mi->seq[i].name, &absent);
if (absent) kh_val(h, k) = i;
else has_dup = 1;
}
mi->h = h;
if (has_dup && mm_verbose >= 2)
fprintf(stderr, "[WARNING] some database sequences have identical sequence names\n");
return has_dup;
}
int mm_idx_name2id(const mm_idx_t *mi, const char *name)