Table 5.
Performance in milliseconds of the one-pass normalization algorithm. The first column is the time required to construct the study model without normalization. The second is the time required for constructing the study model with normalization using the one-pass normalization algorithm. The last two columns show the speedup in milliseconds and as a percentage
Study name | Without de-duplication | With de-duplication | Speedup (ms) | Speedup % |
---|---|---|---|---|
SMALLMR | 556 | 487 | 69 | 12 % |
SMALLCT | 858 | 764 | 94 | 11 % |
TESTMR | 1,270 | 1,199 | 71 | 6 % |
TESTCT | 2,428 | 2,310 | 118 | 5 % |
TESTCTA | 5,114 | 4,600 | 514 | 10 % |
BREASTMR | 5,592 | 5,040 | 552 | 10 % |
Average | 2,636 | 2,400 | 236 | 9 % |