jlview 0.1.0
Initial release.
Features
- Zero-copy R views of Julia Float64 and Int32 arrays via ALTREP
- Automatic type conversion for Float32, Int64, Int16, and UInt8
arrays
- Sparse matrix support (dgCMatrix) with zero-copy nonzero values
- Named vector and matrix support with atomic dimnames (no COW)
- Scope-based release with
with_jlview()
- ALTREP Sum/Min/Max methods calling Julia directly for
performance
- GC pressure tracking with configurable threshold
- Safe cross-runtime garbage collection via C-level finalizers
- Fork safety (mclapply), Julia shutdown safety, double-release
protection
- Full serialization support (saveRDS/readRDS)