aurelius: 'aurelius' package

In aurelius: Generates PFA Documents from R Code and Optionally Runs Them

Description

Generates PFA Documents from R Code and Optionally Runs Them

Details

Converts R syntax into PFA and provides tools for assembling a PFA document within R. Tests validity and runtime behavior of PFA by offloading PFA and data to Titus (through rPython). Facilitates conversion of common R model output to PFA using aurelius.* libraries. Aurelius is part of Hadrian and is on Github at https://github.com/opendatagroup/hadrian.

Examples

## Not run: 
library("aurelius")

# build a model
lm_model <- lm(mpg ~ hp, data = mtcars)
  
# convert the lm object to a list of lists PFA representation
lm_model_as_pfa <- pfa(lm_model)
  
# save as plain-text JSON
write_pfa(lm_model_as_pfa, file = "my-model.pfa")

# read the model back in
read_pfa(file("my-model.pfa"))

## End(Not run)

Example output

$input
$input$type
[1] "record"

$input$fields
$input$fields[[1]]
$input$fields[[1]]$name
[1] "hp"

$input$fields[[1]]$type
[1] "double"



$input$name
[1] "Input"


$output
[1] "double"

$action
$action[[1]]
$action[[1]]$let
$action[[1]]$let$glm_input
$action[[1]]$let$glm_input$type
$action[[1]]$let$glm_input$type$type
[1] "array"

$action[[1]]$let$glm_input$type$items
[1] "double"


$action[[1]]$let$glm_input$new
$action[[1]]$let$glm_input$new[[1]]
[1] "input.hp"





$action[[2]]
$action[[2]]$let
$action[[2]]$let$pred
$action[[2]]$let$pred$model.reg.linear
$action[[2]]$let$pred$model.reg.linear[[1]]
[1] "glm_input"

$action[[2]]$let$pred$model.reg.linear[[2]]
$action[[2]]$let$pred$model.reg.linear[[2]]$cell
[1] "glm_model"






$action[[3]]
[1] "pred"


$cells
$cells$glm_model
$cells$glm_model$type
$cells$glm_model$type$type
[1] "record"

$cells$glm_model$type$fields
$cells$glm_model$type$fields[[1]]
$cells$glm_model$type$fields[[1]]$name
[1] "const"

$cells$glm_model$type$fields[[1]]$type
[1] "double"


$cells$glm_model$type$fields[[2]]
$cells$glm_model$type$fields[[2]]$name
[1] "coeff"

$cells$glm_model$type$fields[[2]]$type
$cells$glm_model$type$fields[[2]]$type$type
[1] "array"

$cells$glm_model$type$fields[[2]]$type$items
[1] "double"




$cells$glm_model$type$name
[1] "gaussianRegression"


$cells$glm_model$init
$cells$glm_model$init$const
[1] 30.09886

$cells$glm_model$init$coeff
$cells$glm_model$init$coeff[[1]]
[1] -0.06822828



$cells$glm_model$source
[1] "embedded"

$cells$glm_model$shared
[1] FALSE

$cells$glm_model$rollback
[1] FALSE

aurelius documentation built on May 2, 2019, 3:43 a.m.