Isoscape Generation
Last Updated: 2023-09-27
Overview
Find the Colab at: https://colab.research.google.com/github/tnc-br/ddf_common/blob/main/generate_isoscape.ipynb
generate_isoscape.ipynb under ddf_common generates isoscapes from a TensorFlow model. This is specifically built to work with the variational inference deep learning model described in Model Training (Variational Inference).
A good place to start is to run the notebook with the default isoscape ("/content/gdrive/MyDrive/amazon_rainforest_files/variational/model/demo_isoscape_model.tf" and "/content/gdrive/MyDrive/amazon_rainforest_files/variational/model/demo_isoscape_model.pkl") and run each cell in order, observing the outputs.
The last cell of code under "Import Tensorflow model and scalers" saves the isoscapes to a geotiff named according to OUTPUT_RASTER_NAME specified at the top of the file. To get the full path, run the following in a Colab cell:
Generated rasters can be imported into the Validation Pipeline.
Last updated
Was this helpful?