Day 15 - apparently much faster with optiimizations?
This commit is contained in:
parent
41d62689df
commit
cfe8ff331c
|
@ -23,7 +23,7 @@ These are the runtimes of only one trial but the variances are fairly small and
|
|||
| 12 | 0.168 |
|
||||
| 13 | 2.136 | 0.503
|
||||
| 14 | 4.360 | 3.495
|
||||
| 15 | 62.242 |
|
||||
| 15 | 62.242 | 3.488
|
||||
| 16 | 0.462 |
|
||||
| 17 | 6.753 | 1.865
|
||||
| 18 | 0.118 |
|
||||
|
@ -31,10 +31,8 @@ These are the runtimes of only one trial but the variances are fairly small and
|
|||
| 20 | 0.168 |
|
||||
| 21 | 4.013 |
|
||||
| 22 | 12.867 | 7.880
|
||||
| 23 | 0.274 |
|
||||
| 23 | 0.274 |
|
||||
|
||||
Problems that should be optimized further: 15
|
||||
|
||||
Problems that could use some work: 05, 22, 14, 21, 17
|
||||
Problems that could use some work: 05, 22, 14, 15, 21, 17
|
||||
|
||||
Problems that are good enough with optimizations: 13
|
||||
|
|
Loading…
Reference in New Issue