Tutorials¶
Guided notebooks teach hrtfpykit through real workflows. They sit between the Quick Start and the API reference: each notebook introduces one public surface, prepares the data it needs, explains the purpose of each step, and shows the code that performs the operation.
Run them interactively in JupyterLab, Jupyter Notebook, or VS Code with Microsoft’s official Jupyter extension. Cell-by-cell execution keeps objects, intermediate values, and generated figures close to the code that produced them.
The first workflows are introductory, but they are still HRTF workflows. It helps to be comfortable running Python notebooks, reading NumPy array shapes, and following core SOFA/HRTF terms such as SOFA files, HRIR, HRTF, source positions, ear channels, samples, frequency bins, and sample rate. Later material builds on that foundation with HRTF transforms, comparison metrics, map-style dataset construction, PyTorch data loaders, batched tensors, model architectures, training loops, validation loops, and HRTF individualization experiments.
Set Up¶
Use a Python environment where hrtfpykit and the notebook tools are
installed. A normal Python environment is enough:
pip install hrtfpykit jupyterlab ipykernel
We recommend using an isolated conda environment.
Create the environment first, activate it, and install hrtfpykit and the
notebook tools inside it:
conda create --name hrtfpykit python
conda activate hrtfpykit
pip install hrtfpykit jupyterlab ipykernel
After installation, open the notebooks with JupyterLab, Jupyter Notebook, or VS
Code. In VS Code, install Microsoft’s official Jupyter extension and select the
environment where hrtfpykit is installed.
If notebooks are new to you, start with the official Jupyter Notebook installation documentation. For details about IPython kernels and managing notebook kernels across Python environments, see the official IPython and IPykernel installation documentation.