SparseWake commited on
Commit
f29f88c
·
verified ·
1 Parent(s): a59067e

Update readme.md

Browse files
Files changed (1) hide show
  1. README.md +69 -3
README.md CHANGED
@@ -1,3 +1,69 @@
1
- ---
2
- license: cc-by-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ pretty_name: SparseWake
4
+ tags:
5
+ - benchmark
6
+ - synthetic-data
7
+ - hydrodynamics
8
+ - time-series
9
+ - regression
10
+ - physical-sensing
11
+ ---
12
+
13
+ # Dataset Name
14
+ SparseWake
15
+
16
+ ## Dataset summary
17
+
18
+ SparseWake evaluates whether sparse body-fixed temporal flow measurements can recover the relative state of a neighboring wake-producing fish in a controlled single-leader setting. The benchmark contains processed synthetic data generated from a DNS-parameterized fish-schooling agent-based model.
19
+
20
+ The released data support the evaluation protocol described in the anonymous submission: held-out follower poses, temporal-history sweeps, sensor-count ablations, component-separated flow inputs, common raw-noise stress tests, and self-signal controls.
21
+
22
+ ## What is included
23
+
24
+ - Processed HDF5 benchmark files.
25
+ - Dataset metadata and sensor coordinates.
26
+ - Held-out-pose split protocol.
27
+ - Summary CSV files for reported tables and figures.
28
+ - Lightweight loading and verification scripts.
29
+ - Checksums and file manifest.
30
+
31
+ ## What is not included
32
+
33
+ The upstream fish-schooling ABM source code and the underlying high-fidelity CFD solver are not redistributed. The release is intended to reproduce benchmark evaluations from the processed HDF5 files, not to regenerate the DNS-parameterized wake library from first principles.
34
+
35
+ ## Intended use
36
+
37
+ This dataset is intended for evaluating sparse temporal physical sensing, hydrodynamic neighbor-state estimation, and reproducibility of the reported benchmark protocol.
38
+
39
+ ## Not Intended Use
40
+
41
+ SparseWake is not a full biological lateral-line pressure model, a multi-neighbor schooling model, or a replacement for direct CFD solvers.
42
+
43
+ ## Data Source
44
+
45
+ The processed data were generated from WakeSchool, an external fish-schooling simulator with DNS-parameterized wake and body-flow components. The artifact starts from processed HDF5 benchmark datasets and does not redistribute simulator source code.
46
+
47
+ ## Released Fields
48
+
49
+ - `X_raw`: wake-plus-potential induced velocity features, sample shape `[N, 6, 3]` after loading.
50
+ - `X_wake_raw`: wake-only induced velocity features where available.
51
+ - `X_potential_raw`: potential-only induced velocity features where available.
52
+ - `X_external_raw`, `X_self_raw`, `X_total_raw`: self-signal controls where available.
53
+ - `y`: labels with columns `delta_x`, `delta_y`, `theta_rel`, `sin_phi`, `cos_phi`, `phi`.
54
+ - `groups`: wake phase index.
55
+ - `pose_id`: pose index for pose-holdout splitting in the sample dataset; full datasets can derive it from phase blocks.
56
+ - `region_id`: region code for close wake, near side, and mid wake.
57
+ - `sensor_world_positions`: sensor coordinates for each sample.
58
+
59
+ ## Units and Conventions
60
+
61
+ Lengths are nondimensionalized by fish body length. Angles are stored in radians. Reported orientation metrics use degrees. The fish body is represented with semi-axes `a = 0.5 L` and `b = 0.075 L`.
62
+
63
+ ## Sensor Layout
64
+
65
+ Six body-fixed sensors are arranged as anterior, midbody, and posterior left/right pairs. Sensor names are stored in documentation and sample HDF5 attributes.
66
+
67
+ ## Ethical Considerations
68
+
69
+ The dataset is synthetic fluid-dynamics data. It contains no human subjects, no animal measurements, and no personal data.