spark_plot_log_manLasso: A SparklyR Logistic Regression Feature Selection Method
In GabeChurch/sparkedatools:

The function compares the AUC between logistic regression models for each feature, dropping each iteratively.

1 2	spark_plot_log_manLasso(sparklyr_table, predictor, num_folds = 3, parallelism = 1)

`sparklyr_table`	is the spark table you will pass to the function. You can pass using a dplyr spark table (tbl).
`predictor`	is the target column to predict
`num_folds`	(default=3) this param passes the number of cross-validation folds to use for each logistic regression model
`parallelism`	(default=1) this param allows us to deploy default models simultaneously

Important package requirements:
You must have ggplot2 installed

Example selection of a spark table and graph
spark_table = tbl(sc, sql("select * from sample_data.iris limit 100"))
outputs = spark_plot_log_manLasso(spark_table, predictor='Species')

GabeChurch/sparkedatools documentation built on June 25, 2019, 12:23 p.m.