lmtest: Interfaces for lmtest package for data science pipelines.

In rbertolusso/intubate: Interface to Popular R Functions for Data Science Pipelines

Description

Interfaces to lmtest functions that can be used in a pipeline implemented by magrittr.

Usage

ntbt_bgtest(data, ...)
ntbt_bptest(data, ...)
ntbt_coxtest(data, ...)
ntbt_dwtest(data, ...)
ntbt_encomptest(data, ...)
ntbt_gqtest(data, ...)
ntbt_grangertest(data, ...)
ntbt_harvtest(data, ...)
ntbt_hmctest(data, ...)
ntbt_jtest(data, ...)
ntbt_raintest(data, ...)
ntbt_resettest(data, ...)

Arguments

data	data frame, tibble, list, ...
...	Other arguments passed to the corresponding interfaced function.

Details

Interfaces call their corresponding interfaced function.

Value

Object returned by interfaced function.

Author(s)

Roberto Bertolusso

Examples

## Not run: 
library(intubate)
library(magrittr)
library(lmtest)

## ntbt_bgtest: Breusch-Godfrey Test for higher-order serial correlation
x <- rep(c(1, -1), 50)
y1 <- 1 + x + rnorm(100)
dta <- data.frame(x, y1)

## or for fourth-order serial correlation
## Original function to interface
bgtest(y1 ~ x, order = 4, data = dta)

## The interface puts data as first parameter
ntbt_bgtest(dta, y1 ~ x, order = 4)

## so it can be used easily in a pipeline.
dta %>%
  ntbt_bgtest(y1 ~ x, order = 4)


## ntbt_bptest: Breusch-Pagan test against heteroskedasticity
## ntbt_gqtest: Goldfeld-Quandt test against heteroskedasticity
## ntbt_hmctest: Harrison-McCabe test for heteroskedasticity
x <- rep(c(-1,1), 50)
err1 <- c(rnorm(50, sd=1), rnorm(50, sd=2))
err2 <- rnorm(100)
y1 <- 1 + x + err1
y2 <- 1 + x + err2
dtah <- data.frame(x, y1, y2)

## Original function to interface
bptest(y1 ~ x, data = dtah)
gqtest(y1 ~ x, data = dtah)
hmctest(y1 ~ x, data = dtah)
bptest(y2 ~ x, data = dtah)
gqtest(y2 ~ x, data = dtah)
hmctest(y2 ~ x, data = dtah)

## The interface puts data as first parameter
ntbt_bptest(dtah, y1 ~ x)
ntbt_gqtest(dtah, y1 ~ x)
ntbt_hmctest(dtah, y1 ~ x)
ntbt_bptest(dtah, y2 ~ x)
ntbt_gqtest(dtah, y2 ~ x)
ntbt_hmctest(dtah, y2 ~ x)

## so it can be used easily in a pipeline.
dtah %>%
  ntbt_bptest(y1 ~ x)
dtah %>%
  ntbt_gqtest(y1 ~ x)
dtah %>%
  ntbt_hmctest(y1 ~ x)
dtah %>%
  ntbt_bptest(y2 ~ x)
dtah %>%
  ntbt_gqtest(y2 ~ x)
dtah %>%
  ntbt_hmctest(y2 ~ x)


## ntbt_coxtest: Cox Test for Comparing Non-Nested Models
## ntbt_encomptest: encompassing test of Davidson & MacKinnon for comparing non-nested models
## ntbt_jtest: Davidson-MacKinnon J test for comparing non-nested models
data(USDistLag)
usdl <- na.contiguous(cbind(USDistLag, lag(USDistLag, k = -1)))
colnames(usdl) <- c("con", "gnp", "con1", "gnp1")

## Original function to interface
coxtest(con ~ gnp + con1, con ~ gnp + gnp1, data = usdl)
encomptest(con ~ gnp + con1, con ~ gnp + gnp1, data = usdl)
jtest(con ~ gnp + con1, con ~ gnp + gnp1, data = usdl)

## The interface puts data as first parameter
ntbt_coxtest(usdl, con ~ gnp + con1, con ~ gnp + gnp1)
ntbt_encomptest(usdl, con ~ gnp + con1, con ~ gnp + gnp1)
ntbt_jtest(usdl, con ~ gnp + con1, con ~ gnp + gnp1)

## so it can be used easily in a pipeline.
usdl %>%
  ntbt_coxtest(con ~ gnp + con1, con ~ gnp + gnp1)
usdl %>%
  ntbt_encomptest(con ~ gnp + con1, con ~ gnp + gnp1)
usdl %>%
  ntbt_jtest(con ~ gnp + con1, con ~ gnp + gnp1)

## ntbt_dwtest: Durbin-Watson test for autocorrelation of disturbances
err1 <- rnorm(100)
x <- rep(c(-1,1), 50)
y1 <- 1 + x + err1
err2 <- filter(err1, 0.9, method="recursive")
y2 <- 1 + x + err2
dta <- data.frame(y1, y2, x)

## Original function to interface
dwtest(y1 ~ x, data = dta)
dwtest(y2 ~ x, data = dta)

## The interface puts data as first parameter
ntbt_dwtest(dta, y1 ~ x)
ntbt_dwtest(dta, y2 ~ x)

## so it can be used easily in a pipeline.
dta %>%
  ntbt_dwtest(y1 ~ x)
dta %>%
  ntbt_dwtest(y2 ~ x)


## ntbt_grangertest: Test for Granger Causality
data(ChickEgg)
## Original function to interface
grangertest(egg ~ chicken, order = 3, data = ChickEgg)
grangertest(chicken ~ egg, order = 3, data = ChickEgg)

## The interface puts data as first parameter
ntbt_grangertest(ChickEgg, egg ~ chicken, order = 3)
ntbt_grangertest(ChickEgg, chicken ~ egg, order = 3)

## so it can be used easily in a pipeline.
ChickEgg %>%
  ntbt_grangertest(egg ~ chicken, order = 3)
ChickEgg %>%
  ntbt_grangertest(chicken ~ egg, order = 3)


## ntbt_harvtest: Harvey-Collier test for linearity
x <- 1:50
y1 <- 1 + x + rnorm(50)
y2 <- y1 + 0.3*x^2
dta <- data.frame(y1, x)

## Original function to interface
harvtest(y1 ~ x, data = dta)

## The interface puts data as first parameter
ntbt_harvtest(dta, y1 ~ x)

## so it can be used easily in a pipeline.
dta %>%
  ntbt_harvtest(y1 ~ x)

## ntbt_raintest: Rainbow test for linearity
x <- c(1:30)
y <- x^2 + rnorm(30,0,2)
dta <- data.frame(x, y)

## Original function to interface
raintest(y ~ x, data = dta)

## The interface puts data as first parameter
ntbt_raintest(dta, y ~ x)

## so it can be used easily in a pipeline.
dta %>%
  ntbt_raintest(y ~ x)


## ntbt_resettest: Ramsey's RESET test for functional form
x <- c(1:30)
y1 <- 1 + x + x^2 + rnorm(30)
y2 <- 1 + x + rnorm(30)
dta <- data.frame(x, y1, y2)

## Original function to interface
resettest(y1 ~ x , power=2, type="regressor", data = dta)
resettest(y2 ~ x , power=2, type="regressor", data = dta)

## The interface puts data as first parameter
ntbt_resettest(dta, y1 ~ x , power=2, type="regressor")
ntbt_resettest(dta, y2 ~ x , power=2, type="regressor")

## so it can be used easily in a pipeline.
dta %>%
  ntbt_resettest(y1 ~ x , power=2, type="regressor")
dta %>%
  ntbt_resettest(y2 ~ x , power=2, type="regressor")

## End(Not run)

rbertolusso/intubate documentation built on May 27, 2019, 3 a.m.