Datasets:
Formats:
parquet
Languages:
English
Size:
1K - 10K
Tags:
remote-sensing
carbon-flux
time-series
earth-system-science
knowledge-guided-machine-learning
process-based-modeling
Delete viewer/README.md
Browse files- viewer/README.md +0 -21
viewer/README.md
DELETED
|
@@ -1,21 +0,0 @@
|
|
| 1 |
-
# Dataset Viewer tables
|
| 2 |
-
|
| 3 |
-
This directory contains lightweight Parquet tables used by the Hugging Face
|
| 4 |
-
Dataset Viewer.
|
| 5 |
-
|
| 6 |
-
Each row corresponds to one sample in an NPZ file and includes:
|
| 7 |
-
|
| 8 |
-
- sample index, split, subset, and source NPZ path
|
| 9 |
-
- per-sample array shapes
|
| 10 |
-
- compact PFT and forest-age summaries
|
| 11 |
-
- the valid-observation fraction for in-situ samples
|
| 12 |
-
|
| 13 |
-
The full multidimensional scientific arrays are not duplicated or flattened
|
| 14 |
-
here. They remain in the NPZ files under `GlobalMask/` and `InSituMatched/`.
|
| 15 |
-
|
| 16 |
-
Regenerate the Viewer tables from the repository root with:
|
| 17 |
-
|
| 18 |
-
```bash
|
| 19 |
-
pip install pandas pyarrow
|
| 20 |
-
python scripts/create_viewer_tables.py --root .
|
| 21 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|