Dependencies

---

Python & OS

LineScope supports the following Python versions:

• Python 3.11
• Python 3.12
• Python 3.13
• Python 3.14
• Python 3.15

And operating systems:

• Linux (Ubuntu, Fedora, etc.)
• Windows
• macOS

Python packages

Required

• click (>=8.4.2)
• psutil (>=5.9)

Optional

Some utilities or configuration options require the installation of additional libraries. Install all the optional dependencies with uv pip install "linescope[full]".

• ipykernel (>=6.29)
• ipython (>=8.20)
• scalene (>=2.3,<2.4)

Spark and Databricks integration reuse PySpark (>=3.5) and the Databricks SDK already provided by the environment. LineScope does not install these runtime libraries through the base package or any extra. Enable Spark profiling with spark=True; startup requires PySpark to be available.

Development

The development dependencies are not installed with the package, and are not required for any of its functionalities. These libraries are only necessary to contribute to the project. Install them with uv sync --locked --all-groups.

Dev

• tox (>=4.25)
• tox-uv (>=1.25)
• uv_build (>=0.11.13,<0.12)

Linting

• databricks-sdk
• pre-commit (>=4.2)
• pre-commit-uv (>=4.1)
• ruff (>=0.11)
• ty (>=0.0.1)

Testing

• ipython (>=8.20)
• nbmake (>=1.5.3)
• pytest (>=8.3)
• pytest-cov (>=6)
• pytest-mock (>=3.14)

Documentation

• mike (>=2.1)
• mkdocs (>=1.6,<2)
• mkdocs-autorefs (>=1.4)
• mkdocs-jupyter (>=0.24.6)
• mkdocs-material (>=9.6)
• mkdocs-simple-hooks (>=0.1.5)
• pymdown-extensions (>=10.14)
• pyyaml (>=6)
• regex (>=2024.11.6)

Utilities

• rust-just (>=1.40)

Demos

• jupyterlab (>=4.4,<5)
• nbconvert (>=7.16,<8)

GPU demos

• torch (>=2.8,<3)

The gpu dependency group installs the CUDA workload library. Install the scalene extra and enable backend="scalene", gpu=True to collect GPU metrics. The default Trace backend does not support device measurements.

Local Spark

• pyspark (>=3.5)

The development spark dependency group supplies PySpark for local integration tests, examples, and documentation notebooks. It is included in dev and needs a compatible Java runtime to execute Spark workloads.