Getting Started¶
Do I need to install anything?¶
No. Every notebook in this course includes an Open in Colab badge at the top. Clicking it opens the notebook directly in Google Colab, which provides a free Python environment with GPU/CPU resources in your browser. The first code cell in each notebook installs any required packages automatically.
If you prefer to run notebooks locally, you will need Python 3.9+ and approximately 20 GB of free disk space for data files. See the prerequisites for details.
Do I need a Planet account or API key?¶
No. The course uses Planet’s Open Data STAC, which provides free access to selected Tanager-1 scenes without an account or API key. The notebooks download data directly from the Open STAC at runtime.
See Accessing Tanager-1 Data via the Open Data STAC for a walkthrough.
What Python background do I need?¶
You should be comfortable with:
Python basics: variables, functions, loops, dictionaries
NumPy: array indexing and basic operations
Matplotlib: creating simple plots
You do not need prior experience with remote sensing, HDF5, or hyperspectral data — the course builds those skills from scratch.
How much memory and storage do I need?¶
RAM: 16 GB minimum. Hyperspectral cubes are large in memory (a single scene at full resolution can exceed 4 GB). Google Colab provides enough RAM for all course exercises.
Storage: ~20 GB free disk space if running locally. In Colab, data is downloaded to the session environment and does not use your local storage.
Data¶
How large are the Tanager-1 data files?¶
A single Tanager-1 .h5 scene (426 bands × ~600 columns × variable rows) typically ranges from 1 GB to several GB depending on scene length. The notebooks download only what is needed for each exercise.
What scenes are used in the course?¶
The course uses publicly available Tanager-1 scenes from Planet’s Open Data STAC. These scenes cover a range of land-cover types and are organized by domain (Agriculture, Snow & Ice, Urban, etc.). For a full description of the scenes, see the Course Datasets page.
Can I use my own Tanager-1 data?¶
Yes. The workflows taught in this course apply to any Tanager-1 scene in HDF-EOS5 format. If you have access to your own scenes (through a Planet subscription or data partnership), you can substitute them at the data-loading step in each notebook.
TanagerSpec¶
What is TanagerSpec?¶
TanagerSpec is Planet Labs’ purpose-built Python package for end-to-end Tanager-1 hyperspectral analysis. It provides a unified interface for loading, preprocessing, visualizing, computing spectral indices, and applying machine learning to Tanager scenes. Module 5 of this course is dedicated to TanagerSpec.
Do I need to complete Modules 1–4 before Module 5?¶
It is strongly recommended. Modules 1–4 build the conceptual and technical foundations that TanagerSpec abstracts. You will understand what TanagerSpec is doing (and why) if you have worked through the earlier modules first. Module 5 is designed as the capstone that ties everything together.
Using the Course¶
Can I use these notebooks for my own research or teaching?¶
Yes. The course is licensed under the Apache License 2.0, which allows free use, sharing, and adaptation with attribution. If you use these materials in a publication or course, please cite the course — see the README for citation instructions.
How do I cite the course?¶
GitHub also provides a Cite this repository button on the repository page (top-right sidebar) that generates APA, MLA, and BibTeX citations from the CITATION.cff file.
I found an error or something unclear — how do I report it?¶
Use the Open issue button in the top-right corner of any page (the GitHub icon menu), or visit the Contributing page for instructions on reporting errors and suggesting fixes.