SDRBench — EXAFEL

X-ray free-electron laser images from the LCLS instrument (SLAC): events x detector panels x rows x columns.

Raw binary arrays, not an Arrow/Parquet dataset. datasets.load_dataset("sdrbench/exafel") 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 EXAFEL 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 484ab4c6a9). metadata/ holds the original SDRBench property and template files. Kept in sync automatically by szcompressor/sdrbench.

• Data provider: LCLS (SLAC)
• Size: 10.4 GB in 3 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("exafel")                # default variant "small"
print(ds.fields)
x = ds["dark"]                     # numpy array, dtype <u2, shape (10, 32, 185, 388)

# 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("exafel", 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/exafel", "10x32x185x388/smd-cxif5315-r129-dark.u16", repo_type="dataset", revision="484ab4c6a935c25c8f5ad09854d71c6c1316912c")
x = np.fromfile(p, dtype="<u2").reshape((10, 32, 185, 388))

Variants and fields

small (default)

sdrbench.dataset("exafel", "small") — folder 10x32x185x388/ — 3 files, 137.8 MB, from SDRBENCH-EXAFEL-10x32x185x388.tar.gz.

10 events of the 32-panel detector as raw counts: raw, dark (background) and calib-fde frames, unsigned 16-bit.

Field	dtype	Shape (C order)	File
dark	<u2	10 x 32 x 185 x 388	10x32x185x388/smd-cxif5315-r129-dark.u16
calib-fde	<u2	10 x 32 x 185 x 388	10x32x185x388/smd-cxif5315-r169-calib-fde.u16
raw	<u2	10 x 32 x 185 x 388	10x32x185x388/smd-cxif5315-r169-raw.u16

large

sdrbench.dataset("exafel", "large") — folder 986x32x185x388/ — 1 files, 9.1 GB, from SDRBENCH-EXAFEL-986x32x185x388.tar.gz.

986 calibrated events of the 32-panel detector, single precision.

Field	dtype	Shape (C order)	File
data	<f4	986 x 32 x 185 x 388	986x32x185x388/EXAFEL-LCLS-986x32x185x388.f32

assembled

sdrbench.dataset("exafel", "assembled") — folder 130x1480x1552/ — 4 files, 1.2 GB, from SDRBENCH-EXAFEL-130x1480x1552.tar.gz.

130 assembled 1480 x 1552 detector images with Bragg peak positions and counts per event.

Field	dtype	Shape (C order)	File
data	<f4	130 x 1480 x 1552	130x1480x1552/SDRBENCH-EXAFEL-data-130x1480x1552.f32
nPeaks	<i8	130	130x1480x1552/SDRBENCH-EXAFEL-nPeaks.i64
peakXPosRaw	<f8	130 x 2048	130x1480x1552/SDRBENCH-EXAFEL-peakXPosRaw-130x2048.d64
peakYPosRaw	<f8	130 x 2048	130x1480x1552/SDRBENCH-EXAFEL-peakYPosRaw-130x2048.d64

Citation

Please cite SDRBench; sdrbench cite exafel 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.