Hugging Face's logo

Datasets:
sdrbench
/
exaalt

Modalities:
Text
Formats:
json
Size:
< 1K
Tags:
scientific-data
lossy-compression
sdrbench
hpc
sz3
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
exaalt / README.md
ayzk's picture
ayzk
Update dataset card and file table
1f26915 verified
preview code
|
Raw History Blame Contribute Delete
6.67 kB
metadata 
license: other
license_name: sdrbench
license_link: https://sdrbench.github.io/
pretty_name: 'SDRBench: EXAALT'
tags:
  - scientific-data
  - lossy-compression
  - sdrbench
  - hpc
  - sz3
configs:
  - config_name: files
    data_files: files.jsonl
    default: true

SDRBench — EXAALT

Molecular dynamics simulation (EXAALT, ECP): atom positions (and velocities for 'small'). Single precision, little-endian.

Raw binary arrays, not an Arrow/Parquet dataset. datasets.load_dataset("sdrbench/exaalt") returns the table of files (files.jsonl: path, dtype, shape, sha256). Read the arrays with the sdrbench package or hf_hub_download + numpy as shown below.

This repository is an unmodified mirror of the EXAALT data of SDRBench, the Scientific Data Reduction Benchmark. The originals are hosted by Argonne National Laboratory on Globus; every archive was unpacked and its files uploaded byte-for-byte (sha256 verified) (revision fdd63c3e68). metadata/ holds the original SDRBench property and template files. Kept in sync automatically by szcompressor/sdrbench.

  • Data provider: EXAALT team (ECP)
  • Size: 4.4 GB in 5 variant(s)
  • Format: raw little-endian binary, C order (slowest dimension first); dtype and shape per field below

Usage 

pip install "sdrbench[sz3]"     # sdrbench + pysz (SZ3)

import numpy as np
import sdrbench
from pysz import sz, szConfig, szErrorBoundMode

ds = sdrbench.dataset("exaalt")                # default variant "small"
print(ds.fields)
x = ds["vx"]                     # numpy array, dtype <f4, shape (2869440,)

# compress with SZ3 (pysz) at a 1e-3 value-range-relative error bound
conf = szConfig()
conf.errorBoundMode = szErrorBoundMode.REL
conf.relErrorBound = 1e-3
compressed, ratio = sz.compress(np.ascontiguousarray(x), conf)
y, _ = sz.decompress(compressed, x.dtype.type, x.shape)
max_err, psnr, nrmse = sz.verify(np.asarray(x), y)
print(f"ratio {ratio:.1f}x, PSNR {psnr:.1f} dB, max error {max_err:.3g}")

Files are fetched on first use (in parallel with ds.download("data/")) and cached, pinned to the revision of the installed sdrbench release; if Hugging Face is unreachable the package falls back to the original archive on Globus. Existing local SDRBench copies can be used with sdrbench.dataset("exaalt", root="/path").

Without the package, with huggingface_hub and numpy only:

from huggingface_hub import hf_hub_download
import numpy as np
p = hf_hub_download("sdrbench/exaalt", "2869440/vx.f32", repo_type="dataset", revision="fdd63c3e68af5a9243c8b954aac0b0a3b0736ad0")
x = np.fromfile(p, dtype="<f4").reshape((2869440,))

Variants and fields

small (default)

sdrbench.dataset("exaalt", "small") — folder 2869440/ — 7 files, 68.9 MB, from SDRBENCH-EXAALT-2869440.tar.gz.

One snapshot of 2,869,440 atoms: positions xx, yy, zz and velocities vx, vy, vz.

Field dtype Shape (C order) File
vx <f4 2869440 2869440/vx.f32
vy <f4 2869440 2869440/vy.f32
vz <f4 2869440 2869440/vz.f32
xx <f4 2869440 2869440/xx.f32
yy <f4 2869440 2869440/yy.f32
zz <f4 2869440 2869440/zz.f32

Other files: 2869440/SDRBENCH-EXAALT-2869440-template_data.txt

copper-1

sdrbench.dataset("exaalt", "copper-1") — folder copper/ — 3 files, 204.1 MB, from SDRBENCH-EXAALT-copper.tar.gz.

Copper system, trajectory 1: positions x, y, z of 3137 atoms over 5423 time steps (time x atom).

Field dtype Shape (C order) File
x <f4 5423 x 3137 copper/dataset1-5423x3137.x.f32.dat
y <f4 5423 x 3137 copper/dataset1-5423x3137.y.f32.dat
z <f4 5423 x 3137 copper/dataset1-5423x3137.z.f32.dat

copper-2

sdrbench.dataset("exaalt", "copper-2") — folder copper/ — 3 files, 1.1 GB, from SDRBENCH-EXAALT-copper.tar.gz.

Copper system, trajectory 2: positions x, y, z of 1077290 atoms over 83 time steps (time x atom).

Field dtype Shape (C order) File
x <f4 83 x 1077290 copper/dataset2-83x1077290.x.f32.dat
y <f4 83 x 1077290 copper/dataset2-83x1077290.y.f32.dat
z <f4 83 x 1077290 copper/dataset2-83x1077290.z.f32.dat

helium-1

sdrbench.dataset("exaalt", "helium-1") — folder helium/ — 3 files, 97.7 MB, from SDRBENCH-EXAALT-helium.tar.gz.

Helium system, trajectory 1: positions x, y, z of 1037 atoms over 7852 time steps (time x atom).

Field dtype Shape (C order) File
x <f4 7852 x 1037 helium/dataset1-7852x1037.x.f32.dat
y <f4 7852 x 1037 helium/dataset1-7852x1037.y.f32.dat
z <f4 7852 x 1037 helium/dataset1-7852x1037.z.f32.dat

helium-2

sdrbench.dataset("exaalt", "helium-2") — folder helium/ — 3 files, 3.0 GB, from SDRBENCH-EXAALT-helium.tar.gz.

Helium system, trajectory 2: positions x, y, z of 106711 atoms over 2338 time steps (time x atom).

Field dtype Shape (C order) File
x <f4 2338 x 106711 helium/dataset2-2338x106711.x.f32.dat
y <f4 2338 x 106711 helium/dataset2-2338x106711.y.f32.dat
z <f4 2338 x 106711 helium/dataset2-2338x106711.z.f32.dat

Citation

Please cite SDRBench; sdrbench cite exaalt prints it:

@inproceedings{zhao2020sdrbench,
  title     = {{SDRBench}: Scientific Data Reduction Benchmark for Lossy Compressors},
  author    = {Zhao, Kai and Di, Sheng and Liang, Xin and Li, Sihuan and Tao, Dingwen and Chen, Zizhong and Cappello, Franck},
  booktitle = {2020 IEEE International Conference on Big Data (Big Data)},
  pages     = {2716--2724},
  year      = {2020}
}

License and terms

The data are distributed by the SDRBench team for research use under the terms of the original data providers listed above (see the SDRBench website); please acknowledge them as requested. Rights remain with the data providers.