fabiencasenave commited on
Commit
e123718
·
verified ·
1 Parent(s): ff4bd87

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +112 -0
README.md CHANGED
@@ -1,4 +1,11 @@
1
  ---
 
 
 
 
 
 
 
2
  dataset_info:
3
  features:
4
  - name: Base_1_2/Zone/GridCoordinates/CoordinateX
@@ -58,3 +65,108 @@ configs:
58
  - split: test
59
  path: data/test-*
60
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: cc-by-sa-4.0
3
+ task_categories:
4
+ - graph-ml
5
+ pretty_name: 2D internal aero CFD RANS dataset, under geometrical variations
6
+ tags:
7
+ - physics learning
8
+ - geometry learning
9
  dataset_info:
10
  features:
11
  - name: Base_1_2/Zone/GridCoordinates/CoordinateX
 
65
  - split: test
66
  path: data/test-*
67
  ---
68
+ <p align='center'>
69
+ <img src='https://i.ibb.co/hJqv4hCt/Logo-VKI-2.png' alt='https://i.ibb.co/hJqv4hCt/Logo-VKI-2.png' width='1000'/>
70
+ <img src='https://i.ibb.co/7NX9z7NQ/image001.png' alt='https://i.ibb.co/7NX9z7NQ/image001.png' width='1000'/>
71
+ </p>
72
+
73
+ ```yaml
74
+ owner: Safran
75
+ license: cc-by-sa-4.0
76
+ data_production:
77
+ type: simulation
78
+ physics: 2D compressible RANS, with Spalart-Allmaras turbulence model
79
+ simulator: Broadcast
80
+ num_samples:
81
+ train: 671
82
+ test: 168
83
+ storage_backend: hf_datasets
84
+
85
+ ```
86
+ This dataset was generated with [`plaid`](https://plaid-lib.readthedocs.io/), we refer to this documentation for additional details on how to extract data from `plaid_sample` objects.
87
+
88
+ The simplest way to use this dataset is to first download it:
89
+ ```python
90
+ from plaid.storage import download_from_hub
91
+
92
+ repo_id = "channel/dataset"
93
+ local_folder = "downloaded_dataset"
94
+
95
+ download_from_hub(repo_id, local_folder)
96
+ ```
97
+
98
+ Then, to iterate over the dataset and instantiate samples:
99
+ ```python
100
+ from plaid.storage import init_from_disk
101
+
102
+ local_folder = "downloaded_dataset"
103
+ split_name = "train"
104
+
105
+ datasetdict, converterdict = init_from_disk(local_folder)
106
+
107
+ dataset = datasetdict[split]
108
+ converter = converterdict[split]
109
+
110
+ for i in range(len(dataset)):
111
+ plaid_sample = converter.to_plaid(dataset, i)
112
+ ```
113
+
114
+ It is possible to stream the data directly:
115
+ ```python
116
+ from plaid.storage import init_streaming_from_hub
117
+
118
+ repo_id = "channel/dataset"
119
+
120
+ datasetdict, converterdict = init_streaming_from_hub(repo_id)
121
+
122
+ dataset = datasetdict[split]
123
+ converter = converterdict[split]
124
+
125
+ for sample_raw in dataset:
126
+ plaid_sample = converter.sample_to_plaid(sample_raw)
127
+ ```
128
+
129
+ Sample features can then be retrieved as follows:
130
+ ```python
131
+ from plaid.storage import load_problem_definitions_from_disk
132
+ local_folder = "downloaded_dataset"
133
+ pb_defs = load_problem_definitions_from_disk(local_folder)
134
+
135
+ # or
136
+ from plaid.storage import load_problem_definitions_from_hub
137
+ repo_id = "channel/dataset"
138
+ pb_defs = load_problem_definitions_from_hub(repo_id)
139
+
140
+
141
+ pb_def = next(iter(pb_defs.values()))
142
+
143
+ plaid_sample = ... # use a method from above to instantiate a plaid sample
144
+
145
+ for t in plaid_sample.get_all_time_values():
146
+ for path in pb_def.input_features:
147
+ feature = plaid_sample.get_feature_by_path(path=path, time=t)
148
+ ...
149
+ for path in pb_def.output_features:
150
+ feature = plaid_sample.get_feature_by_path(path=path, time=t)
151
+ ...
152
+ ```
153
+
154
+ For those familiar with HF's `datasets` library, raw data can be retrieved without using the `plaid` library:
155
+ ```python
156
+ from datasets import load_dataset
157
+
158
+ repo_id = "channel/dataset"
159
+
160
+ datasetdict = load_dataset(repo_id)
161
+
162
+ for split_name, dataset in datasetdict.items():
163
+ for raw_sample in dataset:
164
+ for feat_name in dataset.column_names:
165
+ feature = raw_sample[feat_name]
166
+ ```
167
+ Notice that raw data refers to the variable features only, with a specific encoding for time variable features.
168
+
169
+ ### Dataset Sources
170
+
171
+ - **Papers:**
172
+ - [arxiv](https://arxiv.org/abs/2505.02974)