Closed victor1234 closed 1 year ago
Thanks for the report. Should be fixed, please check with our main branch again.
Thank you! Works for me.
I'm new in this area.
Is there any start guide to how to choose suitable lib? I know my data and hardware but don't know compressed algo terms.
I want to create tool for sparse matrix compression. Means list of uint32_t
sorted/partially sorted indices and list of floats. x86
and arm64
CPUs, crossplatform.
I checked not outdated libs from your list and found
streamvbyte
and FastPFor
. Last one is not support arm64
. So only streamvbyte
, right?
And I still searching for float lossy compression for floats. Could you advice?
The streamvbyte codec is a very good choice, I feel. If it does not meet your needs, we can chat further. But several systems had good luck with streamvbyte.
Closing this issue.
OS: Ubuntu 20.04.5 LTS CPU: Intel Xeon X5660
Step to reproduce:
Output
cmake ..
Output
ctest -V
Output
./perf