Memory Efficient Deconstructed Vectorized Dataframe Interface
Project description
MEDVeDI 
Memory Efficient Deconstructed Vectorized Dataframe Interface.
Design goals:
- Favor performance over nice syntax features. Sacrifice fool-proof for efficient zero-copy operations.
- Ensure ideal micro-performance and optimize for moderate data sizes (megabytes).
- The use-case is API server code that you write once and execute many times.
- Try to stay compatible with the Pandas interface. There is no
Series, however. - Rely on numpy.
- Friends with Arrow.
- Frequently release GIL and depend on native extensions doing unsafe things.
- Test only CPython and Linux.
- Support only x86-64 CPUs with AVX2.
- Support only Python 3.10+.
- 100% test coverage.
Otherwise, you should be way better with regular Pandas.
Medvedi is currently heavily used in production of Athenian.
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
medvedi-0.1.67.tar.gz
(881.3 kB
view details)
File details
Details for the file medvedi-0.1.67.tar.gz.
File metadata
- Download URL: medvedi-0.1.67.tar.gz
- Upload date:
- Size: 881.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.11.2
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 95dffd9456f7b4cadf3c6ed63a4e5957112e7bddbf1ad66e911039305a7f9929 |
|
| MD5 | 5b5b2869f07cd2bcfd8a7a967b20f56b |
|
| BLAKE2b-256 | bd764bded691849b67a411af8e74166c8a35f2d60c5176c68015e00b158db3db |
