mungebits2)

Description Usage Arguments Value

Running a mungepiece achieves the same effect as running the mungebit attached to the mungepiece: the first time it is run, we train the mungebit so it remembers metadata it will need to replicate the operation in a live production setting on a single row of data. The second and subsequent times we run the mungepiece, it will execute the predict function of the underlying mungebit.

1	mungepiece_run(data, ..., `_envir` = parent.frame())

`data`	environment or data.frame. Essentially an environment containing a `data` variable. In this case, that `data` variable will have a side effect enacted on it. If a `data.frame`, then the return value will be the modified `data.frame` and the mungebit will record any results it must memorize in its `input`.
`...`	additional arguments to the mungebit's `train_function` or `predict_function`.
`_envir`	environment. The calling environment for the train or predict function on the underlying mungebit. This is an internal argument and is `parent.frame()` by default.

If the data parameter is an environment, the transformed environment (i.e., the transformed data in the environment) after application of the underlying mungebit. If data is a data.frame, the transformed data.frame is returned.

syberia/mungebits2 documentation built on May 30, 2019, 10:41 p.m.