icebug

Icebug - Frozen Graph Analytics

Icebug is an open source library for high-performance graph analysis backed by read-only memory. Due to its heritage, it also supports network/graph analysis on read-write memory, but it can be 5x more CPU and 5x more memory efficient when using columnar memory such as Apache Arrow.

Its aim is to provide tools for the analysis of large networks in the size range from thousands to billions of edges. For this purpose, it implements efficient graph algorithms, many of them parallel to utilize multicore architectures. These are meant to compute standard measures of network analysis. Icebug is focused on scalability and comprehensiveness. Icebug is also a testbed for algorithm engineering and contains novel algorithms from recently published research (see list of publications below).

Icebug is a Python module. High-performance algorithms are written in C++ and exposed to Python via the Cython toolchain. Python in turn gives us the ability to work interactively and a rich environment of tools for data analysis and scientific computing. Furthermore, Icebug's core can be built and used as a native library if needed.

Requirements

You will need the following software to install Icebug as a python package:

A modern C++ compiler, e.g.: g++ (>= 8.1), clang++ (>= 6.0) or MSVC (>= 14.20)
OpenMP for parallelism (usually ships with the compiler)
Python3 (3.9 or higher is supported)
- Development libraries for Python3. The package name depends on your distribution. Examples:
  - Debian/Ubuntu: apt-get install python3-dev
  - RHEL/CentOS: dnf install python3-devel
  - Windows: Use the official release installer from www.python.org
Pip
CMake version 3.6 or higher (Advised to use system packages if available. Alternative: pip3 install cmake)
Build system: Make or Ninja
Cython version 0.29 or higher (e.g., pip3 install cython)
Apache Arrow development libraries (libarrow-dev)

Install

In order to use Icebug, you can either install it via package managers or build the Python module from source.

Install via package manager

While the most recent version is in general available for all package managers, the number of older downloadable versions differ.

uv

uv pip install icebug

More system-specific information on how to install Icebug on Linux, macOS (both Intel and M1) and Windows-systems can be found here.

Building the Python module from source

git clone https://github.com/Ladybug-Memory/icebug icebug
cd icebug
python3 setup.py build_ext [-jX]
pip3 install -e .

The script will call cmake and ninja (make as fallback) to compile Icebug as a library, build the extensions and copy it to the top folder. By default, Icebug will be built with the amount of available cores in optimized mode. It is possible the add the option -jN the number of threads used for compilation.

Usage example

To get an overview and learn about Icebug's different functions/classes, have a look at our interactive notebooks-section, especially the Icebug UserGuide. Note: To view and edit the computed output from the notebooks, it is recommended to use Jupyter Notebook. This requires the prior installation of Icebug. You should really check that out before start working on your network analysis.

We also provide a Binder-instance of our notebooks. To access this service, you can either click on the badge at the top or follow this link. Disclaimer: Due to rebuilds of the underlying image, it can takes some time until your Binder instance is ready for usage.

If you only want to see in short how Icebug is used - the following example provides a glimpse at that. Here we generate a random hyperbolic graph with 100k nodes and compute its communities with the PLM method:

>>> import networkit as nk
>>> g = nk.generators.HyperbolicGenerator(1e5).generate()
>>> communities = nk.community.detectCommunities(g, inspect=True)
PLM(balanced,pc,turbo) detected communities in 0.14577102661132812 [s]
solution properties:
-------------------  -----------
# communities        4536
min community size      1
max community size   2790
avg. community size    22.0459
modularity              0.987243
-------------------  -----------

Icebug also offers two ways to load a graph from CSR format enabling zero-copy loading of large graphs using Arrow tables

>>> import pyarrow as pa
>>> from icebug_format import IcebugMemGraph
>>> nodes = pa.table({"id": pa.array([0, 1, 2], type=pa.int64())})
>>> rels  = pa.table({"source": pa.array([0, 1, 2], type=pa.int64()),
...                   "target": pa.array([1, 2, 0], type=pa.int64())})
>>> mem = IcebugMemGraph.from_arrow_tables(nodes, rels)
>>> g = nk.graph.Graph.fromIcebugMemGraph(mem)

# using fromCSR directly
>>> indices = pa.table({"target": pa.array([1, 2, 0], type=pa.int64())})
>>> indptr = pa.table({"row_ptr": pa.array([0, 1, 2, 3], type=pa.int64())})
>>> g = nk.graph.Graph.fromCSR(3, True, indices, indptr)

Note: Properties in IcebugMemGraph are pyarrow tables, so, node and edge tables could contain properties, whereas the fromCSR method only supports loading the graph structure, so no properties are supported. If the resolved CSR column is already a single uint64 chunk, fromIcebugMemGraph keeps it zero-copy; chunked columns are combined and integer columns are cast to uint64, which allocates.

Install the C++ Core only

In case you only want to work with Icebug's C++ core, you can either install it via package managers or build it from source.

Install C++ core via package manager

conda (channel conda-forge)

conda config --add channels conda-forge
conda install libicebug [-c conda-forge]

brew

brew install libicebug

spack

spack install libicebug

Building the C++ core from source

We recommend CMake and your preferred build system for building the C++ part of Icebug.

The following description shows how to use CMake in order to build the C++ Core only:

First you have to create and change to a build directory: (in this case named build)

mkdir build
cd build

Then call CMake to generate files for the make build system, specifying the directory of the root CMakeLists.txt file (e.g., ..). After this make is called to start the build process:

cmake ..
make -jX

To speed up the compilation with make a multi-core machine, you can append -jX where X denotes the number of threads to compile with.

Use NetworKit as a library

This paragraph explains how to use the Icebug core C++ library in case it has been built from source. For how to use it when installed via package managers, best refer to the official documentation (brew, conda, spack).

In order to use the previous compiled icebug library, you need to have it installed, and link it while compiling your project. Use these instructions to compile and install Icebug in /usr/local:

cmake ..
make -jX install

Once Icebug has been installed, you can use include directives in your C++-application as follows:

#include <icebug/graph/Graph.hpp>

You can compile your source as follows:

g++ my_file.cpp -licebug

Unit tests

Building and running Icebug unit tests is not mandatory. However, as a developer you might want to write and run unit tests for your code, or if you experience any issues with Icebug, you might want to check if Icebug runs properly. The unit tests can only be run from a clone or copy of the repository and not from a pip installation. In order to run the unit tests, you need to compile them first. This is done by setting the CMake NETWORKI_BUILD_TESTS flag to ON:

cmake -DNETWORKIT_BUILD_TESTS=ON ..

Unit tests are implemented using GTest macros such as TEST_F(CentralityGTest, testBetweennessCentrality). Single tests can be executed with:

./networkit_tests --gtest_filter=CentralityGTest.testBetweennessCentrality

Additionally, one can specify the level of the logs outputs by adding --loglevel <log_level>; supported log levels are: TRACE, DEBUG, INFO, WARN, ERROR, and FATAL.

Compiling with address/leak sanitizers

Sanitizers are great tools to debug your code. Icebug provides additional Cmake flags to enable address, leak, and undefined behavior sanitizers. To compile your code with sanitizers, set the CMake NETWORKIT_WITH_SANITIZERS to either address or leak:

cmake -DNETWORKIT_WITH_SANITIZERS=leak ..

By setting this flag to address, your code will be compiled with the address and the undefined sanitizers. Setting it to leak also adds the leak sanitizer.

Documentation

The most recent version of the documentation can be found online.

Contact

For questions regarding Icebug, have a look at our issues-section and see if there is already an open discussion. If not feel free to open a new issue. To stay updated about this project, subscribe to our mailing list.

Contributions

We encourage contributions to the Icebug source code. See the development guide for instructions. For support please contact the mailing list.

Credits

Icebug is a fork of NetworKit.

List of contributors can be found on the NetworKit website credits page.

External Code

The program source includes:

the TLX library
the TTMath bignum library

License

The source code of this program is released under the MIT License. We ask you to cite us if you use this code in your project (c.f. the publications section below and especially the technical report). Feedback is also welcome.

Publications

The Icebug publications page lists the publications on Icebug as a toolkit, on algorithms available in Icebug, and simply using Icebug. We ask you to cite the appropriate ones if you found Icebug useful for your own research.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

icebug-13.0.tar.gz (7.5 MB view details)

Uploaded Jul 27, 2026 Source

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

icebug-13.0-cp314-cp314-win_amd64.whl (36.2 MB view details)

Uploaded Jul 27, 2026 CPython 3.14Windows x86-64

icebug-13.0-cp314-cp314-manylinux_2_28_x86_64.whl (28.2 MB view details)

Uploaded Jul 27, 2026 CPython 3.14manylinux: glibc 2.28+ x86-64

icebug-13.0-cp314-cp314-manylinux_2_28_aarch64.whl (26.7 MB view details)

Uploaded Jul 27, 2026 CPython 3.14manylinux: glibc 2.28+ ARM64

icebug-13.0-cp314-cp314-macosx_15_0_x86_64.whl (20.2 MB view details)

Uploaded Jul 27, 2026 CPython 3.14macOS 15.0+ x86-64

icebug-13.0-cp314-cp314-macosx_15_0_arm64.whl (19.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.14macOS 15.0+ ARM64

icebug-13.0-cp313-cp313-win_amd64.whl (35.7 MB view details)

Uploaded Jul 27, 2026 CPython 3.13Windows x86-64

icebug-13.0-cp313-cp313-manylinux_2_28_x86_64.whl (28.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.13manylinux: glibc 2.28+ x86-64

icebug-13.0-cp313-cp313-manylinux_2_28_aarch64.whl (26.5 MB view details)

Uploaded Jul 27, 2026 CPython 3.13manylinux: glibc 2.28+ ARM64

icebug-13.0-cp313-cp313-macosx_15_0_x86_64.whl (20.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.13macOS 15.0+ x86-64

icebug-13.0-cp313-cp313-macosx_15_0_arm64.whl (19.0 MB view details)

Uploaded Jul 27, 2026 CPython 3.13macOS 15.0+ ARM64

icebug-13.0-cp312-cp312-win_amd64.whl (35.7 MB view details)

Uploaded Jul 27, 2026 CPython 3.12Windows x86-64

icebug-13.0-cp312-cp312-manylinux_2_28_x86_64.whl (28.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.12manylinux: glibc 2.28+ x86-64

icebug-13.0-cp312-cp312-manylinux_2_28_aarch64.whl (26.5 MB view details)

Uploaded Jul 27, 2026 CPython 3.12manylinux: glibc 2.28+ ARM64

icebug-13.0-cp312-cp312-macosx_15_0_x86_64.whl (20.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.12macOS 15.0+ x86-64

icebug-13.0-cp312-cp312-macosx_15_0_arm64.whl (19.0 MB view details)

Uploaded Jul 27, 2026 CPython 3.12macOS 15.0+ ARM64

icebug-13.0-cp311-cp311-win_amd64.whl (35.7 MB view details)

Uploaded Jul 27, 2026 CPython 3.11Windows x86-64

icebug-13.0-cp311-cp311-manylinux_2_28_x86_64.whl (28.2 MB view details)

Uploaded Jul 27, 2026 CPython 3.11manylinux: glibc 2.28+ x86-64

icebug-13.0-cp311-cp311-manylinux_2_28_aarch64.whl (26.6 MB view details)

Uploaded Jul 27, 2026 CPython 3.11manylinux: glibc 2.28+ ARM64

icebug-13.0-cp311-cp311-macosx_15_0_x86_64.whl (20.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.11macOS 15.0+ x86-64

icebug-13.0-cp311-cp311-macosx_15_0_arm64.whl (19.0 MB view details)

Uploaded Jul 27, 2026 CPython 3.11macOS 15.0+ ARM64

icebug-13.0-cp310-cp310-win_amd64.whl (35.7 MB view details)

Uploaded Jul 27, 2026 CPython 3.10Windows x86-64

icebug-13.0-cp310-cp310-manylinux_2_28_x86_64.whl (28.2 MB view details)

Uploaded Jul 27, 2026 CPython 3.10manylinux: glibc 2.28+ x86-64

icebug-13.0-cp310-cp310-manylinux_2_28_aarch64.whl (26.6 MB view details)

Uploaded Jul 27, 2026 CPython 3.10manylinux: glibc 2.28+ ARM64

icebug-13.0-cp310-cp310-macosx_15_0_x86_64.whl (20.1 MB view details)

Uploaded Jul 27, 2026 CPython 3.10macOS 15.0+ x86-64

icebug-13.0-cp310-cp310-macosx_15_0_arm64.whl (19.0 MB view details)

Uploaded Jul 27, 2026 CPython 3.10macOS 15.0+ ARM64

File details

Details for the file icebug-13.0.tar.gz.

File metadata

Download URL: icebug-13.0.tar.gz
Upload date: Jul 27, 2026
Size: 7.5 MB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.14

File hashes

Hashes for icebug-13.0.tar.gz
Algorithm	Hash digest
SHA256	`d956ce424ceaa6cbc383e44ff85786553d543289daa5ef9fe4a4525ed89418c3`
MD5	`7efe3977c4dc4a35e31498153fc2c5fa`
BLAKE2b-256	`051171fd7a1ebbacf22b965de868bf3a657574a7337ea506ca6e172fafb1d296`

Algorithm	Hash digest
SHA256	`1561655621c1adac96894daa353760ce1189a8d024db69fbfba981a04e94e3a9`
MD5	`588323cf8472b199d06b6ad96b74460c`
BLAKE2b-256	`d28eca7440eaa64194525ada595cc07aa2b5fa3bc83322d060f42bfa372b648b`

Algorithm	Hash digest
SHA256	`e9712fc0a482dbf336d549a4fa0874a2a90a3b07b6e68216a63311fd12ef775c`
MD5	`260cb65b3c911022602559e5d35266c8`
BLAKE2b-256	`1afc0bf683f8ce217be2b740fac3e7dcfeda23ec3a79767df7cfe925f34d8a0a`

Algorithm	Hash digest
SHA256	`b815a02fc85cdc678777f272feb73a9160139c5ce800ec809d3117e5d883c55c`
MD5	`b6668e1837d1640a84b66155d7107a3f`
BLAKE2b-256	`5a61602ae8d3497c99c274867bf87ebf4b800668dc0e5b2a934eded6564c4516`

Algorithm	Hash digest
SHA256	`1387ef6db4dd5a04d35ba8f744004b9fe9d172567fb3b6eb2609562dd67915bf`
MD5	`8c8b5934fe8ab8f73c564a17d3e35f2a`
BLAKE2b-256	`ece979cd98f2a55e9ce01688bdb64bbabbe76a262fc26e94c0743c9639291514`

Algorithm	Hash digest
SHA256	`4770b8e8177d5109debfd0389995714efc0959081bfc95df0ee8b71da30116e5`
MD5	`cde43dcda235f0bc455946d0c1d06523`
BLAKE2b-256	`7b53de942556b0bdb2e1511c6a553b2208311071a16dbf5806beb6eab9e7bd96`

Algorithm	Hash digest
SHA256	`788e226865be301885a8d8e3ed957d762c2a185dce53ac0e4d202ea44b23e646`
MD5	`8e20a20289d98a95cc917b37568bd0d0`
BLAKE2b-256	`83dcb7d7e3fe678094990803f240ee43daf5376768f69828f4caa0f81e8b2a42`

Algorithm	Hash digest
SHA256	`2f9a1a37f520069abd3b8f878d1c7abd20b16a1d8c15fb179c0b0094a727ff27`
MD5	`d60928f48ed6aa6c63edc0c4b2624674`
BLAKE2b-256	`9ff506ce1eb29af04548d544393675f1aa3bf3765ebd3e46832e57c77c700555`

Algorithm	Hash digest
SHA256	`f05f93a96c29bc1750a2d45513331a0dd814aceaf9effaef291f4b7a25d43fb4`
MD5	`f69a8ea4bde952ec0841138224b10f4a`
BLAKE2b-256	`af999039ab57cc66558fe5049969d13f04944fbeacfc49b5fc0da5e6d2ae1882`

Algorithm	Hash digest
SHA256	`baec5fac8c18ca6ba1b1cf84d1cb17551f2a899541dc8fe092df0a5770bd48e5`
MD5	`b692fe8a5ec5ff31f6bc9445cfc3a3e2`
BLAKE2b-256	`15f237e44d491cb1514df1ba395497ecaf34f5cf68bc197f953175e2792ae9cd`

Algorithm	Hash digest
SHA256	`470c9d7795f35128f7a72ffa40c9db87e1dbc3ecadb88faeeb9882107f416eb0`
MD5	`35742c213fed8e11af10ad0ea4dbed40`
BLAKE2b-256	`7125778e977997f0c57a56bcde882a90e6d20dd4ef180c51d166db8c99745b4e`

Algorithm	Hash digest
SHA256	`037953e0f4e75a38f6932a1951e1e89359d99832b50ce514cc2df8ce66b1cf45`
MD5	`ad5d525cec4a056ae3d9b7ccbbcab5fb`
BLAKE2b-256	`a2ec47eb5f3b59d9c3167e95ec3cf10ff6aff2813462ca9774b6b00acafd7693`

Algorithm	Hash digest
SHA256	`4dfc38c54fa5070fd6359bd846f294c7e48d740078a22942e2a6380edd3b4288`
MD5	`456b1eadb0d376c752317b761a9ffce6`
BLAKE2b-256	`9228683412ec0ade290e8437add608bdd2c961a98ef088a5e8a5a7deed60bf2e`

Algorithm	Hash digest
SHA256	`2335370c8d821b94c4bb328445345212abb879a28bd452b728e3926627454186`
MD5	`e7c37e7de91d04a2bef686de3244f84b`
BLAKE2b-256	`5e4e3beb327e0474f83d19607a60407b2dcb89137253372448b2be619a5aa80a`

Algorithm	Hash digest
SHA256	`c049549d034396df43d8b48bb2449e4ff126aced87fcb62f15eddaa3fc8f3bc5`
MD5	`9536ca58880fd047283e0895d8a11068`
BLAKE2b-256	`8c47266933fdfd7de861afa75a21aa74015d6876061db0ed1a333dbd3866affc`

Algorithm	Hash digest
SHA256	`6a0cef88f784db22e29ebb7735c9f710a8b13e9cb6918c61b1a19aed91503525`
MD5	`ec36fe2de86b96dff8ac495ee796784a`
BLAKE2b-256	`339db7254b98c0fad2f8b4d5816b312e03927dfdf04745f6cf45d4b69f6af3a6`

Algorithm	Hash digest
SHA256	`28d27818b7bbc6f60d9cca013baa8547bddefae8ed4c7517dba7d9b184c430e5`
MD5	`5ebf255bb4d893ff4e16c1457a535ec1`
BLAKE2b-256	`0d4e9e76fa53755daaf33dd82a98723cc32c67ae8f31ccf7bceabb6145b6ca41`

Algorithm	Hash digest
SHA256	`ba87c73ac4ff4a006da0dd51ae9a57aad25b5893bd2f426927b69712d6bf1f8f`
MD5	`e4202bea539b6f0770a97e49f86d4a39`
BLAKE2b-256	`4a5ccde253f070d2aed58d7eca1da153134fd92dbb09de104dc8609d6d287587`

Algorithm	Hash digest
SHA256	`320bc969d5b2248dcacabc0816494852bc5b9d5061fb0588f90c851d823937a9`
MD5	`d07e8a5e24ea3be54c60800460ddc083`
BLAKE2b-256	`c4bb8e74e3a34738b02b03998159a76099dc1afb33df44a2d8191faf1c3100da`

Algorithm	Hash digest
SHA256	`c6d2e26b189238001c718111ed950940144a9c7542ef23027d2b48fba8632977`
MD5	`1efcc20b43ddd1e8f873cbd9cc58ff3b`
BLAKE2b-256	`5c63550c77cf620015828ff71c5f92498851beae1eb8f3dd7d718931eb445ab8`

Algorithm	Hash digest
SHA256	`4611dbfaeb5f02e6cff559c54210b66bdc2717e1383123dd401860570167385d`
MD5	`8dc03261186e45020032b0aff2ae8380`
BLAKE2b-256	`886edebf84b4d44407f6faf8f56d51bd0fa8b5b5838fc3676c12e462aac97d39`

icebug 13.0

Requirements

Install

Install via package manager

uv

Building the Python module from source

Usage example

Install the C++ Core only

Install C++ core via package manager

conda (channel conda-forge)

brew

spack

Building the C++ core from source

Use NetworKit as a library

Unit tests

Compiling with address/leak sanitizers

Documentation

Contact

Contributions

Credits

External Code

License

Publications

Project links

Key dates

1 maintainer

Credits

GitHub Statistics

License

Requires

Provides Extra

Tags

Classifiers

Download files

Source Distribution

Built Distributions

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

Algorithm	Hash digest
SHA256	`1f4f19196c656b8e4c946f2e9fdaf96521a62e45512a3536111bf88a74a868f3`
MD5	`411cdc22e858f8e33ba3b19503ab1e42`
BLAKE2b-256	`cce97d2035fc33968d87825f1cf2ff00e1aaac76b709e13956a8764c8edd2609`

Algorithm	Hash digest
SHA256	`ed5edb1e56926c9d5475af69ef5d0fb40af92fd8534d43aa6f942b6df2772b2a`
MD5	`dbbb31714b11ecbb2ffc1e0c38050605`
BLAKE2b-256	`8e63c8e88547fa138c20a9e7f2dd466ac9834cf3799cacf27fc3e19eedac44f1`

Algorithm	Hash digest
SHA256	`9f05674601c480cda31ce270b70b79226d1f6419a6198f765d6a3523a3df358e`
MD5	`02e6ed4890adee20e16623681fc509e9`
BLAKE2b-256	`448f38e19aca3e3c4c34d3f802a5de10a78ce15b2b874257ab7b4df9270730c6`

Algorithm	Hash digest
SHA256	`6851e7a74b0d3661491529f172766f6e5e28a26808bd0387c96b6bfc500d9058`
MD5	`84e770e832aa41ce1cd4bb8f189d603e`
BLAKE2b-256	`ac92c49b40ac06b0ec99cf2fdd8b4609eb42bcc7685011027ea7e51a9bc05ffe`

Algorithm	Hash digest
SHA256	`085d5eafb1378d6eb702c2186aeebdf6ea810f332b68778d33b5d36dbfadcb78`
MD5	`9aa0f6150598923f234c13a61d29c9e4`
BLAKE2b-256	`a9172243b4d24e5bc343e4fb1e2ff924f10b63d432daf9ebcb59b43707f13273`