aliwister commited on
Commit
5976b64
·
verified ·
1 Parent(s): 51fe72c

Add geese and stork trajectory datasets with dataset card

Browse files
README.md ADDED
@@ -0,0 +1,138 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc0-1.0
3
+ pretty_name: WildGraph Wildlife Migration Trajectories
4
+ task_categories:
5
+ - other
6
+ tags:
7
+ - trajectory-generation
8
+ - trajectory
9
+ - wildlife
10
+ - animal-movement
11
+ - migration
12
+ - gps
13
+ - geospatial
14
+ - movebank
15
+ size_categories:
16
+ - 10K<n<100K
17
+ configs:
18
+ - config_name: geese
19
+ data_files:
20
+ - split: train
21
+ path: geese/train-00000-of-00001.parquet
22
+ - config_name: stork
23
+ data_files:
24
+ - split: train
25
+ path: stork/train-00000-of-00001.parquet
26
+ ---
27
+
28
+ # WildGraph: wildlife migration trajectories
29
+
30
+ The two bird migration datasets used in
31
+
32
+ > Ali Al-Lawati, Elsayed Eshra, Prasenjit Mitra.
33
+ > [WildGraph: Realistic Long-Horizon Trajectory Generation with Limited Sample Size](https://doi.org/10.1145/3678717.3691265).
34
+ > SIGSPATIAL 2024. ([arXiv:2404.08068](https://arxiv.org/abs/2404.08068))
35
+
36
+ Each dataset is a small set of months-long GPS trajectories with one location per day and the same length for
37
+ every trajectory. They were prepared for long-horizon trajectory generation from few samples: a model sees a
38
+ few dozen real migrations and has to generate new, realistic ones.
39
+
40
+ ## Contents
41
+
42
+ Two configurations, each with a single `train` split and one row per daily location.
43
+
44
+ | Configuration | Species | Clipping window | Days per trajectory | Trajectories | Rows |
45
+ | --- | --- | --- | ---: | ---: | ---: |
46
+ | `geese` | Greater white-fronted goose (*Anser albifrons*) | Mar 1 to Sep 1 | 185 | 60 | 11,100 |
47
+ | `stork` | White stork (*Ciconia ciconia*) | Aug 15 to Nov 1 | 79 | 77 | 6,083 |
48
+
49
+ The geese trajectories lie between 48.5° and 76.2° N and between 2.8° and 113.2° E, and the stork trajectories
50
+ between 9.3° and 53.0° N and between 6.4° W and 37.7° E.
51
+
52
+ ```python
53
+ from datasets import load_dataset
54
+
55
+ ds = load_dataset("aliwister/WildGraph", "geese", split="train")
56
+ print(ds[0])
57
+ # {'idx': 9065, 'label': 0, 'day': 0, 'location.lat': 52.0339758, 'location.long': 6.1129115}
58
+
59
+ # all trajectories as one array of shape (trajectories, days, 2), ordered (latitude, longitude)
60
+ df = ds.to_pandas()
61
+ trajs = df[["location.lat", "location.long"]].to_numpy().reshape(df["label"].nunique(), -1, 2)
62
+ print(trajs.shape)
63
+ # (60, 185, 2)
64
+ ```
65
+
66
+ ## Fields
67
+
68
+ | Field | Type | Description |
69
+ | --- | --- | --- |
70
+ | `idx` | int64 | Row identifier kept from the released CSV files. It is unique within a configuration and consecutive within a trajectory, but it does not order the trajectories. |
71
+ | `label` | int64 | Trajectory identifier, 0 to 59 for `geese` and 0 to 76 for `stork`. One label is one animal in one season (see below). |
72
+ | `day` | int64 | Position of the location within its trajectory, starting at 0. Day 0 corresponds to the first day of the clipping window. |
73
+ | `location.lat` | float64 | Latitude in decimal degrees (WGS84). |
74
+ | `location.long` | float64 | Longitude in decimal degrees (WGS84). |
75
+
76
+ Rows are stored trajectory by trajectory and in day order. The files hold no timestamps, years, or animal
77
+ identifiers.
78
+
79
+ ## How the data was built
80
+
81
+ Both datasets were obtained from the [Movebank Data Repository](https://www.movebank.org) and processed as
82
+ described in Section 5.4 of the paper.
83
+
84
+ 1. Observations were reduced to one per day for each animal with [MoveApps](https://www.moveapps.org).
85
+ 2. Each track was clipped to the window in the table above, which covers one migration.
86
+ 3. Animals tracked over several years were split into one trajectory per year, so a `label` is an animal in
87
+ one season, and two labels can belong to the same animal.
88
+ 4. Trajectories with significant missing readings were dropped, and the remaining isolated gaps were filled by
89
+ simple interpolation. A small number of locations are therefore interpolated and not measured.
90
+
91
+ ## Use
92
+
93
+ The paper trains on the trajectories of four folds and compares generated trajectories with the held-out fold,
94
+ using five-fold cross-validation over `label` without shuffling.
95
+
96
+ ```python
97
+ from sklearn.model_selection import KFold
98
+
99
+ labels = range(df["label"].nunique())
100
+ for k, (train_labels, test_labels) in enumerate(KFold(n_splits=5).split(labels)):
101
+ train, test = df[df["label"].isin(train_labels)], df[df["label"].isin(test_labels)]
102
+ ```
103
+
104
+ Generated trajectories are scored by their distance to the real ones (Hausdorff distance, dynamic time warping,
105
+ and final displacement error), by how many held-out trajectories they cover, and by how closely their spatial
106
+ distribution matches the real one (Pearson correlation and chi-squared similarity). Code for WildGraph, the
107
+ baselines, and the metrics is at [github.com/aliwister/wildgraph](https://github.com/aliwister/wildgraph). The
108
+ code reads the columns `idx`, `label`, `location.lat`, and `location.long`, which keep their original names
109
+ here, so `ds.to_pandas()` can replace the CSV files of that repository.
110
+
111
+ ## Source data and license
112
+
113
+ The trajectories are derived from two studies published in the Movebank Data Repository under CC0 1.0. This
114
+ dataset is released under the same terms. Please cite the source data and the papers they accompany when you
115
+ use it.
116
+
117
+ - `geese`: Kölzsch A, Müskens GJDM, Moonen S, Kruckenberg H, Glazov P, Wikelski M (2019). Data from: Flyway
118
+ connectivity and exchange primarily driven by moult migration in geese [North Sea population]. Movebank Data
119
+ Repository. [doi:10.5441/001/1.ct72m82n](https://doi.org/10.5441/001/1.ct72m82n). Accompanies Kölzsch et al.
120
+ (2019), Movement Ecology 7:3, [doi:10.1186/s40462-019-0148-6](https://doi.org/10.1186/s40462-019-0148-6).
121
+ - `stork`: Carlson B, Rotics S, Nathan R, Wikelski M, Jetz W (2021). Data from: Individual environmental niches
122
+ in mobile organisms. Movebank Data Repository.
123
+ [doi:10.5441/001/1.rj21g1p1](https://doi.org/10.5441/001/1.rj21g1p1). Accompanies Carlson et al. (2021),
124
+ Nature Communications 12:4572, [doi:10.1038/s41467-021-24826-x](https://doi.org/10.1038/s41467-021-24826-x).
125
+
126
+ ## Citation
127
+
128
+ ```bibtex
129
+ @inproceedings{allawati2024wildgraph,
130
+ title = {WildGraph: Realistic Long-Horizon Trajectory Generation with Limited Sample Size},
131
+ author = {Al-Lawati, Ali and Eshra, Elsayed and Mitra, Prasenjit},
132
+ booktitle = {Proceedings of the 32nd ACM International Conference on Advances in Geographic Information Systems},
133
+ publisher = {ACM},
134
+ pages = {247--258},
135
+ year = {2024},
136
+ doi = {10.1145/3678717.3691265}
137
+ }
138
+ ```
geese/train-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0e343bd248dc954399887299fc2cb1fe5936923278f0e8a782b74085d57ecb0
3
+ size 283807
stork/train-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e6d74c70c5fb1aae626cf7f39f59b3542bb7da8efb37ec6eda0c29f5d195816
3
+ size 156684