Predict#
Used to make predictions
- class predict.predict(X_train, y, X_predict, model_config)#
- Parameters:
X_train ({array-like, sparse matrix} of shape (n_samples, n_features)) – should be the same as the X used for tuning
y (array-like of shape (n_samples,) or (n_samples, n_outputs)) – should be the same as the y used for tuning
X_predict ({array-like, sparse matrix} of shape (n_samples, n_features)) – Features to predict on (i.e. gridded environmental data).
model_config (dictionary, default=None) –
A dictionary containing:
seed : int, used to create random numbers
root: string, path to folder
path_out: string, where predictions are saved
path_in: string, where to find tuned models
traits: string, file name of your trait file
verbose: int, to set verbosity (0-3)
n_threads: int, number of threads to use
cv : int, number of cross-folds
ensemble_config :
clf_scoring :
reg_scoring :
- make_prediction()#
Calculates performance of model(s) and exports prediction(s) to netcdf
Notes
If more than one model is provided, predictions are made for both invidiual models and an ensemble of the models.