def sketch_scalar_product(M, c): result_sketch = CMSketch(M.m, M.d) for i in range(M.d): for j in range(M.m): new_val = c*M.val_at(i, j) result_sketch.update(i, j, new_val) return result_sketch
def _make_sketch(kmer_counts_dict: defaultdict) -> CountMinSketch:
# Read the dictionary into a compressed data structure to allow deleting kmer_counts_dict
NUM_ROWS = 8
kmer_counts = CountMinSketch(NUM_ROWS)
for kmer, count in kmer_counts_dict.items():
kmer_counts.update(kmer, count)
return kmer_counts
def sketch_sum(M1, M2): if M1.m != M2.m: print "Sketches don't align on hashtable length.\n" return elif M1.d != M2.d: print "Sketches don't align on # of hashtables.\n" return else: result_sketch = CMSketch(M2.m, M2.d) for i in range(M2.d): for j in range(M2.m): new_val = M1.val_at(i, j) + M2.val_at(i, j) result_sketch.update(i, j, new_val) return result_sketch
def _make_sketch(self, kmer_counts_dict: defaultdict) -> CountMinSketch:
if self.print_runtime:
print("\n>--- STARTING TO MAKE COUNTMIN SKETCH AT T = {:.2f} ---".
format(time.time() - self.start_time))
# Read the dictionary into a compressed data structure
NUM_ROWS = 10
kmer_counts = CountMinSketch(NUM_ROWS)
for i, (kmer, count) in enumerate(kmer_counts_dict.items()):
if self.print_runtime and i % 50000 == 0:
print(">Processed {0} kmers by time T={1:.2f}".format(
i,
time.time() - self.start_time))
kmer_counts.update(kmer, count)
if self.print_runtime:
print(">FINISHED MAKING COUNTMIN SKETCH AT T = {:.2f}".format(
time.time() - self.start_time))
if self.print_syssizeof:
print(">SIZE OF COUNTMIN SKETCH: {:,}".format(
sys.getsizeof(kmer_counts)))
return kmer_counts
