pipe_component: Default (do-nothing) pipeline component

In kota7/MLPipe: Machine Learning Pipeline

Description

This is a pipeline component doing nothing. It is mainly used like a virtual class, where functions are defined in children classes.

Usage

pipe_component(...)

Arguments

...

not in use

Value

PipeComponent class object

Class Attributes

objects

an R object. Typically used to store a model object

not_parameters

character vector that specifies the names of fields not considered as parameters

Class Methods

fit(x=NULL, y=NULL)

do nothing

incr_fit(x=NULL, y=NULL)

do nothing

transpose(x=NULL, y=NULL)

return x and y as-is

inv_transform(x=NULL, y=NULL)

return x and y as-is

predict(x=NULL, ...)

return NULL

predict_proba(x=NULL, ...)

return NULL

set_parameters(...)

change parameter values, cannot create new fields

get_parameters(param_names=character(0)

return parameters as list. if param_names are the names is null or length 0, return all parameters

get_parameter_namess()

return the names of all parameters

Examples

p <- pipe_component()
p$fit(x=c(1,5,8), y=c('a','b','a'))  # nothing happens
p$transform(x=1:10, y=1:10)          # return data as-is
p$get_parameters()                   # parameters as list
p$get_parameter_names()              # parameter names

kota7/MLPipe documentation built on May 5, 2019, 5:53 p.m.