h2o_predict: Calculate predictions of h2o Models

In lares: Lean Analytics and Robust Exploration Sidekick

Calculate predictions of h2o Models

Description

h2o_predict_MOJO lets the user predict using the h2o .zip file containing the MOJO files. Note that it works with the files generated when using the function export_results()

h2o_predict_binary lets the user predict using the h2o binary file. Note that it works with the files generated when using the function export_results(). Recommendation: use the h2o_predict_MOJO() function when possible - it let's you change h2o's version without problem.

h2o_predict_model lets the user get scores from a H2O Model Object.

h2o_predict_API lets the user get the score from an API service

Usage

h2o_predict_MOJO(df, model_path, method = "mojo", batch = 300)

h2o_predict_binary(df, model_path, sample = NA)

h2o_predict_model(df, model)

h2o_predict_API(df, api, exclude = "tag")

Arguments

df	Dataframe/Vector. Data to insert into the model.
model_path	Character. Relative model path directory or zip file.
method	Character. One of "mojo" or "json".
batch	Integer. Run n batches at a time for "json" method.
sample	Integer. How many rows should the function predict?
model	h2o model Object
api	Character. API URL.
exclude	Character. Name of the variables to exclude.

Value

data.frame with predicted results.

vector with predicted results.

data.frame with predicted results.

vector with predicted results.

See Also

Other Machine Learning: ROC(), conf_mat(), export_results(), gain_lift(), h2o_automl(), h2o_selectmodel(), impute(), iter_seeds(), lasso_vars(), model_metrics(), model_preprocess(), msplit()

lares documentation built on Aug. 18, 2025, 5:12 p.m.