hypso.experimental.chlorophyll.estimate_chlorophyll_ml module

hypso.experimental.chlorophyll.estimate_chlorophyll_ml.get_best_features(X_train_rrs, y_train_rrs, X_test_rrs, y_test_rrs, dataset_name='hypso')

Compute the optimal features based on the findings of the Master thesis “Improving Ocean Chlorophyll Estimation in Satellite Hyperspectral Images Using Ensemble Machine Learning” by Flores-Romero Alvaro

Parameters:

• X_train_rrs (ndarray) – Surface reflectance values for training

• y_train_rrs (ndarray) – Surface chlorophyll values for training

• X_test_rrs (ndarray) – Surface reflectance values for testing

• y_test_rrs (ndarray) – Surface chlorophyll values for testing

• dataset_name (str) – Dataset name to estimate values. Defaults to “hypso”

Returns:

Returns the features for training and testing

Return type:

Tuple[DataFrame, DataFrame]

hypso.experimental.chlorophyll.estimate_chlorophyll_ml.start_chl_estimation(sat_obj, model_path=None)

Estimates the chlorophyll using a trained ML Scikit-Learn Model

Parameters:

• sat_obj – Hypso satellite object

• model_path – Absolute path for the pre-trained Scikit-Learn model

Returns:

No return.

Return type:

None