Package: mizu 0.1.1.9000

mizu: Thread-Speed Async and Parallel R over Shared Memory

Parallel computation and data exchange between R processes on the same machine: lock-free channels, work-stealing task pools, and a parallel map over 'POSIX' shared memory (Linux, macOS) or 'Win32' file mappings (Windows), with the hot path entirely in user space. Communication is cheap enough that work can be divided as finely as with threads. Vectors and data frames transfer zero-copy, received as 'ALTREP' views over the shared pages, and 'Apache Arrow' tables and record batches travel as a single frame with zero-copy conversion in both directions. Receives and task results integrate with 'later' and 'promises' for non-blocking, event-driven applications such as 'shiny'. Channels, pools, and maps interoperate with 'pymizu', so a peer or worker can equally be a Python process. It complements 'mirai', and its payloads interoperate with 'mori' shared objects.

Authors:Charlie Gao [aut, cre], Posit Software, PBC [cph, fnd], Pierre L'Ecuyer [cph]

mizu_0.1.1.9000.tar.gz
mizu_0.1.1.9000.zip(r-4.7-x86_64)mizu_0.1.1.9000.zip(r-4.7-arm64)mizu_0.1.1.9000.zip(r-4.6-x86_64)mizu_0.1.1.9000.zip(r-4.6-arm64)mizu_0.1.1.9000.zip(r-4.5-x86_64)
mizu_0.1.1.9000.tgz(r-4.6-x86_64)mizu_0.1.1.9000.tgz(r-4.6-arm64)mizu_0.1.1.9000.tgz(r-4.5-x86_64)mizu_0.1.1.9000.tgz(r-4.5-arm64)
mizu_0.1.1.9000.tar.gz(r-4.7-arm64)mizu_0.1.1.9000.tar.gz(r-4.7-x86_64)mizu_0.1.1.9000.tar.gz(r-4.6-arm64)mizu_0.1.1.9000.tar.gz(r-4.6-x86_64)
mizu_0.1.1.9000.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
mizu/json (API)

# Install 'mizu' in R:
install.packages('mizu', repos = c('https://shikokuchuo.r-universe.dev', 'https://cloud.r-project.org'))

Bug tracker:https://github.com/shikokuchuo/mizu/issues

Pkgdown/docs site:https://shikokuchuo.net

On CRAN:

Conda:

asynchronous-programmingshared-memory-parallelismquarto

8.13 score 50 stars 48 scripts 518 downloads 45 exports 0 dependencies

Last updated from:593ce8b1c7. Checks:4 ERROR, 11 OK. Indexed: yes.

TargetResultTimeFilesSyslog
linux-devel-arm64ERROR203
linux-devel-x86_64ERROR187
source / vignettesOK201
linux-release-arm64ERROR186
linux-release-x86_64ERROR151
macos-release-arm64OK187
macos-release-x86_64OK328
macos-oldrel-arm64OK194
macos-oldrel-x86_64OK272
windows-devel-arm64OK486
windows-devel-x86_64OK587
windows-release-arm64OK567
windows-release-x86_64OK302
windows-oldrel-x86_64OK204
wasm-releaseOK88

Exports:mizu_alivemizu_as_arrowmizu_callmizu_cancelmizu_channelmizu_channel_infomizu_closemizu_collectmizu_collect_allmizu_collect_all_asyncmizu_collect_anymizu_collect_any_asyncmizu_collect_asyncmizu_current_poolmizu_default_poolmizu_is_remote_errormizu_is_sentinelmizu_launchermizu_local_poolmizu_mapmizu_map_preparemizu_map_runmizu_poolmizu_pool_attachmizu_pool_dumpmizu_pool_infomizu_pool_statsmizu_pool_stopmizu_pool_tracemizu_prunemizu_py_launchermizu_recvmizu_recv_asyncmizu_recv_batchmizu_recv_batch_asyncmizu_recv_handlermizu_retire_workermizu_sendmizu_send_batchmizu_set_default_poolmizu_spawn_workersmizu_submitmizu_submit_batchmizu_submit_callmizu_with_pool

Dependencies:

Channels
Promises and handlers
Rendered fromchannels.qmdusingquarto::html

Last update: 2026-11-03
Started: 2026-10-01

Parallel map
Templates | Reproducible randomness | Prepared maps | Nested maps
Rendered frommap.qmdusingquarto::html

Last update: 2026-11-03
Started: 2026-10-01

Python interop
Large data crosses without copying | What crosses | What doesn’t cross | Data frames and Arrow | Mixed-language pools | The cross-language map
Rendered frominterop.qmdusingquarto::html

Last update: 2026-11-03
Started: 2026-10-01

Task pools
Nested tasks | A default pool for package code | Waiting on several tasks | Promises and handlers | Sizing, sharing and watching a pool
Rendered frompools.qmdusingquarto::html

Last update: 2026-11-03
Started: 2026-10-01

Benchmarks
A data frame round trip | Five thousand small tasks | How mizu compares | Base R | mirai, crew, and future | callr | Python workers
Rendered frombenchmarks.qmdusingquarto::html

Last update: 2026-11-03
Started: 2026-10-01

Introduction to mizu
The articles
Rendered frommizu.qmdusingquarto::html

Last update: 2026-11-02
Started: 2026-10-01

Operations
Deploying on Linux | Linux memory allocator | Garbage collection boundary | When a process dies
Rendered fromoperations.qmdusingquarto::html

Last update: 2026-11-02
Started: 2026-10-01

Readme and manuals

Help Manual

Help pageTopics
Check Whether the Channel Peer Is Alivemizu_alive
Convert a Received Frame to an Arrow Streammizu_as_arrow
Call a Function in Another Languagemizu_call mizu_submit_call
Cancel a Task or a Mapmizu_cancel
Create a Channel and Spawn Its Peermizu_channel
Inspect a Channelmizu_channel_info
Close a Channelmizu_close
Collect the Results of Several Tasks, in Ordermizu_collect_all
Collect Pool Results as Promisesas.promise.mizu_map as.promise.mizu_task is.promising.mizu_map is.promising.mizu_task mizu_collect_all_async mizu_collect_any_async mizu_collect_async
Collect the First Available Result From Several Tasksmizu_collect_any
The Evaluating Worker's Own Pool Handlemizu_current_pool
The Process-Wide Default Poolmizu_default_pool mizu_local_pool mizu_set_default_pool mizu_with_pool
Error Conditionsmizu_error mizu_error_cancelled mizu_error_not_portable mizu_error_python_payload mizu_error_remote mizu_error_shm mizu_error_slots_exhausted mizu_error_startup mizu_error_stopped mizu_error_submit_timeout mizu_error_task_lost mizu_error_worker_died
Remote Errorsmizu_is_remote_error
Test for a mizu Sentinelmizu_is_sentinel
Default Child Process Launchermizu_launcher
Parallel Map Over a Poolmizu_map
Prepared Maps: Stage Once, Run Manymizu_map_prepare mizu_map_run
Create a Task Pool and Spawn Its Workersmizu_pool
Attach to a Pool as a Submittermizu_pool_attach
Dump the Distributed State of a Poolmizu_pool_dump
Inspect a Poolmizu_pool_info
Cumulative Pool Countersmizu_pool_stats
Stop a Poolmizu_pool_stop
Trace Task Lifecycle Eventsmizu_pool_trace
Remove Orphaned Shared Memory Regionsmizu_prune
Python Launchermizu_py_launcher
Receive Channel Messages as a Promise or a Streamas.promise.mizu_channel is.promising.mizu_channel mizu_recv_async mizu_recv_batch_async mizu_recv_handler
Send and Receive over a Channelmizu_recv mizu_send
Batched Send and Receivemizu_recv_batch mizu_send_batch
Grow or Shrink the Worker Set of a Poolmizu_retire_worker mizu_spawn_workers
Submit a Task and Collect Its Resultmizu_collect mizu_submit
Submit a Batch of Tasksmizu_submit_batch
Print Methods for mizu Objectsprint.mizu_call print.mizu_channel print.mizu_error_remote print.mizu_map print.mizu_map_prepared print.mizu_pool print.mizu_sentinel print.mizu_task