Release 1.4.13.3

support time-dimension based compression in addition to single snapshot compression
./configure --prefix=[Your installation path] --enable-timecmpr
make && make install

You will find testfloat_compress_ts.c and testfloat_decompress_ts.c in the example/ directory.

testing data can be downloaded from here:
http://www.mcs.anl.gov/~shdi/download/consecutive-steps.tar.gz

Release 1.4.13.2 (support openMP + clean-up of codes)

Please see README for details.
The performance and compression quality should be the same as 1.4.12/13.0

Release 1.4.13.0 (support openMP)

1. Support openMP
2. Integrate PaSTRI algorithm into SZ for GAMESS two-electron integral data compression.

Release a new version 1.4.12.3 for adding a few more ease-of-use options

1. Add -M to specify the error controlling mode, and -A/-R/-P to specify the specific error bounds.
2. Add -p to print the meta data based on a given compressed data file.
3. Allowing users to specify the output file path for compression data file or decompressed data file. If not specified, the compressed data file is *.sz and the decompressed data file is *.sz.out by default.
4. Add hands-on document in doc/, which is more helpful for beginners.
5. Add some checkings in the executable command sz, to print informative messages when users make mistakes about the input such as wrong configuration file path or input file path.
6. Remove the dependency of sz.config for sz command.

A new release with a few fixed bugs and more options in executable

Fix a bug happening when the error bound is extremely small.
Fix a bug in SZ_Init_Params()
Add more options to support error controls in the executable command (see sz -h for details)
Revised user guide to respond to the new options in the executable command.

Release a new version 1.4.12.1

SZ 1.4.12: (1) Support thresholding-based strategy for 1D data compression based on point-wise relative error bound. (In order to test it, please select errBoundMode = PW_REL, and set the point-wise relative error bound using the parameter pw_relBoundRatio in the sz.config.) For other dimensions of data, point-wise relative error based compression is using block-based strategy (see our DRBSD-2 paper for details) (2) fix the bug in the callZlib.c (previously, segmentation fault might happen when using best_compression mode). (2) Fix a small bug that happened when the data size is extremely huge (nbEle>4G) and the compression mode is SZ_BEST_COMPRSSION. Specifically, the previous call to zlib functions has one potential bug that may lead to segmentation fault, which has been fixed.
SZ 1.4.12.1: (1) Fix the bug tha segmentation fault may happen when the error bound is greater than the value range size.

Release 1.4.11.1

In this new release, we support all types of integer compression, also in HDF5-SZ

New release 1.4.11.0 (in Sept. 2017)

(1) Support HDF5. (2) Support integer data compression. (3) Provide optional wavelet transform as a preprocessing step in SZ and an optional Tucker tensor decomposition.

Release SZ-1.4.10.0

(1) Support HDF5 filter;
(2) Support direct sub-block data compression;
(3) Support compression of large data file directly (i.e., the number of data points could be up to as large as LONG size, unlike the previous version that can only compress 2^{32} data points each time);
(4) Separate the internel functions from the sz.h;
(5) Supports more type of error modes.

v1.4.9.2

SZ 1.4.9-beta w/ complete user guide.