Window options (ADR-0005 input)¶
Generated by uv run python scripts/window_options.py from the cached slot counts only (ADR-0004 slots, placeholder-dated BODMAS benign rows excluded). No features were loaded and no model was trained. below target counts slots whose raw malware share is already under the ratio (cap and flag: they keep all rows).
LAMDA (month slots)¶
Usable runs at ratio 0.10, M = 50¶
first |
last |
n_slots |
n_below_target |
min_malware_kept |
|---|---|---|---|---|
month:2013-06 |
month:2013-06 |
1 |
0 |
2611 |
month:2013-09 |
month:2014-08 |
12 |
0 |
178 |
month:2016-02 |
month:2016-04 |
3 |
1 |
497 |
month:2016-06 |
month:2016-06 |
1 |
0 |
50 |
month:2017-01 |
month:2017-04 |
4 |
1 |
342 |
month:2017-10 |
month:2018-04 |
7 |
0 |
109 |
month:2018-06 |
month:2018-08 |
3 |
0 |
411 |
month:2018-11 |
month:2023-12 |
62 |
31 |
103 |
month:2024-02 |
month:2024-03 |
2 |
1 |
55 |
month:2024-08 |
month:2024-09 |
2 |
2 |
55 |
Usable slots by target ratio (M = 50)¶
ratio |
usable_slots |
below_target |
longest_run |
runs |
|---|---|---|---|---|
0.05 |
95 |
8 |
47 |
10 |
0.10 |
97 |
36 |
62 |
10 |
0.20 |
97 |
50 |
62 |
10 |
Usable slots by size gate (ratio 0.10)¶
min_malware |
usable_slots |
longest_run |
|---|---|---|
30 |
101 |
66 |
50 |
97 |
62 |
100 |
93 |
62 |
Windows per candidate (ratio 0.10, M = 50)¶
nT |
nE |
stride |
n_windows |
n_eval_slots_distinct |
eval_reuse |
|---|---|---|---|---|---|
3 |
6 |
1 |
58 |
68 |
5.12 |
3 |
6 |
3 |
20 |
66 |
1.82 |
3 |
6 |
6 |
10 |
60 |
1.00 |
3 |
12 |
1 |
48 |
59 |
9.76 |
3 |
12 |
3 |
16 |
57 |
3.37 |
3 |
12 |
6 |
8 |
54 |
1.78 |
4 |
6 |
1 |
56 |
66 |
5.09 |
4 |
6 |
3 |
19 |
63 |
1.81 |
4 |
6 |
6 |
10 |
60 |
1.00 |
4 |
12 |
1 |
47 |
58 |
9.72 |
4 |
12 |
3 |
16 |
57 |
3.37 |
4 |
12 |
6 |
8 |
54 |
1.78 |
BODMAS (2w slots)¶
Usable runs at ratio 0.10, M = 50¶
first |
last |
n_slots |
n_below_target |
min_malware_kept |
|---|---|---|---|---|
2w:2019-08-26 |
2w:2019-12-16 |
9 |
0 |
136 |
2w:2020-02-10 |
2w:2020-09-21 |
17 |
0 |
133 |
Usable slots by target ratio (M = 50)¶
ratio |
usable_slots |
below_target |
longest_run |
runs |
|---|---|---|---|---|
0.05 |
26 |
0 |
17 |
2 |
0.10 |
26 |
0 |
17 |
2 |
0.20 |
27 |
1 |
17 |
2 |
Usable slots by size gate (ratio 0.10)¶
min_malware |
usable_slots |
longest_run |
|---|---|---|
30 |
27 |
17 |
50 |
26 |
17 |
100 |
26 |
17 |
Windows per candidate (ratio 0.10, M = 50)¶
nT |
nE |
stride |
n_windows |
n_eval_slots_distinct |
eval_reuse |
|---|---|---|---|---|---|
3 |
4 |
1 |
14 |
20 |
2.80 |
3 |
4 |
2 |
8 |
20 |
1.60 |
3 |
4 |
3 |
5 |
17 |
1.18 |
3 |
5 |
1 |
12 |
20 |
3.00 |
3 |
5 |
2 |
6 |
18 |
1.67 |
3 |
5 |
3 |
5 |
19 |
1.32 |
3 |
6 |
1 |
10 |
20 |
3.00 |
3 |
6 |
2 |
6 |
20 |
1.80 |
3 |
6 |
3 |
4 |
18 |
1.33 |
3 |
8 |
1 |
7 |
14 |
4.00 |
3 |
8 |
2 |
4 |
14 |
2.29 |
3 |
8 |
3 |
3 |
14 |
1.71 |
4 |
4 |
1 |
12 |
18 |
2.67 |
4 |
4 |
2 |
6 |
16 |
1.50 |
4 |
4 |
3 |
5 |
17 |
1.18 |
4 |
5 |
1 |
10 |
18 |
2.78 |
4 |
5 |
2 |
6 |
18 |
1.67 |
4 |
5 |
3 |
4 |
16 |
1.25 |
4 |
6 |
1 |
8 |
13 |
3.69 |
4 |
6 |
2 |
4 |
12 |
2.00 |
4 |
6 |
3 |
3 |
12 |
1.50 |
4 |
8 |
1 |
6 |
13 |
3.69 |
4 |
8 |
2 |
3 |
12 |
2.00 |
4 |
8 |
3 |
2 |
11 |
1.45 |