Get Started¶

Heliokit targets a controlled 64-bit CPython 3.10 environment. The current packages are pinned to Python 3.10 because the desktop repository contains compiled extensions and native scientific dependencies built for that runtime.

Choose a Python 3.10 environment¶

There are two supported installation paths. The Conda path is recommended: it keeps Heliokit and its native scientific dependencies isolated from other Python projects.

Use an existing Python 3.10 environment

Use a managed 64-bit CPython 3.10 installation that already has the required runtime dependencies. Confirm the interpreter before installing:
```
python --version
python -c "import struct; print(struct.calcsize('P') * 8)"
```
The commands must report Python 3.10.x and 64 respectively.
Create a Conda Python 3.10 environment (recommended)

Create and activate a dedicated environment, then update pip inside that environment:
```
conda create --name heliokit python=3.10 --yes
conda activate heliokit
python -m pip install --upgrade pip
```
Run all later python and pip commands from this activated environment. A different environment name is fine, provided it is a 64-bit Python 3.10 environment.

Install Heliokit¶

Install Heliokit from PyPI:

python -m pip install heliokit

This installs the host package and the standard first-party data, filter, tool, and visualization components. It intentionally excludes model components and the geomagnetic forecast pipeline because those packages pull in TensorFlow- or Torch-related dependencies.

The equivalent explicit spelling is:

python -m pip install "heliokit[no-model]"

Install model components only when you need DST, Kp, SEP, TVD-MHD, Transformer, server-dynamic models, the MCCHD CNN workflow, or the geomagnetic pipeline:

python -m pip install "heliokit[models]"

Use the all-inclusive form when you want every standard component plus the model stack:

python -m pip install "heliokit[all]"

Inspect components¶

heliokit components list
heliokit components list --kind data

Run a component¶

heliokit data read vti D:\data\volume.vti

histogram --input D:\data\sample.plt --array BR --save D:\plots\histogram.png --no-show

Use installed shortcuts¶

pip install heliokit installs shortcut commands for first-party components into the active Python environment. After conda activate heliokit, commands such as xy_line, histogram, volume, and meshtovtk_scs can be run directly:

xy_line --help
vti D:\data\volume.vti
histogram --input D:\data\sample.plt --array BR --save D:\plots\histogram.png --no-show

Use heliokit activate only when you want optional PowerShell prompt integration that shows (heliokit).