pROC: Interfaces for pROC package for data science pipelines.

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

Description

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

Usage

ntbt_auc(data, ...)
ntbt_ci(data, ...)
ntbt_ci.auc(data, ...)
ntbt_ci.coords(data, ...)
ntbt_ci.se(data, ...)
ntbt_ci.sp(data, ...)
ntbt_ci.thresholds(data, ...)
ntbt_multiclass.roc(data, ...)
ntbt_plot.roc(data, ...)
ntbt_roc(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(pROC)

## NOTE: pROC examples below use both formula and non-formula variants.
##       In examples for other packages, almost always only
##       the formula variant is shown, but in those cases also
##       the non-formula variants should work.

## ntbt_auc: Compute the area under the ROC curve
data(aSAH)

## Original function to interface
auc(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
auc(outcome, s100b)
detach()
## or use $
auc(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
## NOTE: in this case the formula version fails, and I have found no
##       way to trick auc into accepting the formula (so far).
##       Maybe (only maybe) there is a problem with auc, as formula
##       variant may not be used, so it was probably not
##       reported as a bug before. The rest of the interfaced
##       functions seem to work fine.
## ntbt_auc(data = aSAH, outcome ~ s100baSAH)
## The non-formula variant works fine
ntbt_auc(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
#aSAH %>%
#  ntbt_auc(outcome ~ s100baSAH)
aSAH %>%
  ntbt_auc(outcome, s100b)


## ntbt_ci: Compute the confidence interval of a ROC curve
## Original function to interface
ci(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
ci(outcome, s100b)
detach()
## or use $
ci(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
ntbt_ci(aSAH, outcome ~ s100b)
ntbt_ci(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
aSAH %>%
  ntbt_ci(outcome ~ s100b)
aSAH %>%
  ntbt_ci(outcome, s100b)


## ci.auc: Compute the confidence interval of the AUC
## Original function to interface
ci.auc(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
ci.auc(outcome, s100b)
detach()
## or use $
ci.auc(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
ntbt_ci.auc(aSAH, outcome ~ s100b)
ntbt_ci.auc(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
aSAH %>%
  ntbt_ci.auc(outcome ~ s100b)
aSAH %>%
  ntbt_ci.auc(outcome, s100b)


## ntbt_ci.coords: Compute the confidence interval of arbitrary coordinates
## Original function to interface
set.seed(1)
ci.coords(outcome ~ s100b, data = aSAH, x="best", input = "threshold", 
          ret=c("specificity", "ppv", "tp"))
set.seed(1)
ci.coords(aSAH$outcome, aSAH$s100b, x="best", input = "threshold", 
          ret=c("specificity", "ppv", "tp"))
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
set.seed(1)
ci.coords(outcome, s100b, x="best", input = "threshold", 
          ret=c("specificity", "ppv", "tp"))
detach()
## or use $
set.seed(1)
ci.coords(aSAH$outcome, aSAH$s100b, x="best", input = "threshold", 
          ret=c("specificity", "ppv", "tp"))

## The interface puts data as first parameter
set.seed(1)
ntbt_ci.coords(aSAH, outcome ~ s100b, x="best", input = "threshold", 
               ret=c("specificity", "ppv", "tp"))
set.seed(1)
ntbt_ci.coords(aSAH, outcome, s100b, x="best", input = "threshold", 
               ret=c("specificity", "ppv", "tp"))

## so it can be used easily in a pipeline.
set.seed(1)
aSAH %>%
  ntbt_ci.coords(outcome ~ s100b, x="best", input = "threshold", 
                 ret=c("specificity", "ppv", "tp"))
set.seed(1)
aSAH %>%
  ntbt_ci.coords(outcome, s100b, x="best", input = "threshold", 
                 ret=c("specificity", "ppv", "tp"))


## ntbt_ci.se: Compute the confidence interval of sensitivities at given specificities
## Original function to interface
set.seed(1)
ci.se(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
set.seed(1)
ci.se(outcome, s100b)
detach()
## or use $
set.seed(1)
ci.se(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
set.seed(1)
ntbt_ci.se(aSAH, outcome ~ s100b)
set.seed(1)
ntbt_ci.se(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
set.seed(1)
aSAH %>%
  ntbt_ci.se(outcome ~ s100b)
set.seed(1)
aSAH %>%
  ntbt_ci.se(outcome, s100b)


## ntbt_ci.sp: Compute the confidence interval of specificities at given sensitivities
## Original function to interface
set.seed(1)
ci.sp(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
set.seed(1)
ci.sp(outcome, s100b)
detach()
## or use $
set.seed(1)
ci.sp(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
set.seed(1)
ntbt_ci.sp(aSAH, outcome ~ s100b)
set.seed(1)
ntbt_ci.sp(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
set.seed(1)
aSAH %>%
  ntbt_ci.sp(outcome ~ s100b, x="best", input = "threshold", 
             ret=c("specificity", "ppv", "tp"))
set.seed(1)
aSAH %>%
  ntbt_ci.sp(outcome, s100b, x="best", input = "threshold", 
             ret=c("specificity", "ppv", "tp"))


## ntbt_ci.thresholds: Compute the confidence interval of thresholds
## Original function to interface
set.seed(1)
ci.thresholds(outcome ~ s100b, data = aSAH)
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
set.seed(1)
ci.thresholds(outcome, s100b)
detach()
## or use $
set.seed(1)
ci.thresholds(aSAH$outcome, aSAH$s100b)

## The interface puts data as first parameter
set.seed(1)
ntbt_ci.thresholds(aSAH, outcome ~ s100b)
set.seed(1)
ntbt_ci.thresholds(aSAH, outcome, s100b)

## so it can be used easily in a pipeline.
set.seed(1)
aSAH %>%
  ntbt_ci.thresholds(outcome ~ s100b)
set.seed(1)
aSAH %>%
  ntbt_ci.thresholds(outcome, s100b)


## ntbt_multiclass.roc: Multi-clmulticlass.roc Multi-class AUCass AUC
## Original function to interface
multiclass.roc(gos6 ~ s100b, data = aSAH, levels = c(3, 4, 5))
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
multiclass.roc(gos6, s100b, levels = c(3, 4, 5))
detach()
## or use $
multiclass.roc(aSAH$gos6, aSAH$s100b, levels = c(3, 4, 5))

## The interface puts data as first parameter
ntbt_multiclass.roc(aSAH, gos6 ~ s100b, levels = c(3, 4, 5))
ntbt_multiclass.roc(aSAH, gos6, s100b, levels = c(3, 4, 5))

## so it can be used easily in a pipeline.
aSAH %>%
  ntbt_multiclass.roc(gos6 ~ s100b, levels = c(3, 4, 5))
aSAH %>%
  ntbt_multiclass.roc(gos6, s100b, levels = c(3, 4, 5))


## ntbt_plot.roc: Plot a ROC curve
## Original function to interface
plot.roc(outcome ~ s100b, data = aSAH, type="b", pch=21, col="blue", bg="grey")
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
plot.roc(outcome, s100b, type="b", pch=21, col="blue", bg="grey")
detach()
## or use $
plot.roc(aSAH$outcome, aSAH$s100b, type="b", pch=21, col="blue", bg="grey")

## The interface puts data as first parameter
ntbt_plot.roc(aSAH, outcome ~ s100b, type="b", pch=21, col="blue", bg="grey")
ntbt_plot.roc(aSAH, outcome, s100b, type="b", pch=21, col="blue", bg="grey")

## so it can be used easily in a pipeline.
aSAH %>%
  ntbt_plot.roc(outcome ~ s100b, type="b", pch=21, col="blue", bg="grey")
aSAH %>%
  ntbt_plot.roc(outcome, s100b, type="b", pch=21, col="blue", bg="grey")


## ntbt_roc: Build a ROC curve
## Original function to interface
roc(outcome ~ s100b, data = aSAH, type="b", pch=21, col="blue", bg="grey")
## For non-formula variants, either:
## 1) need to attach
attach(aSAH)
roc(outcome, s100b, type="b", pch=21, col="blue", bg="grey")
detach()
## or use $
roc(aSAH$outcome, aSAH$s100b, type="b", pch=21, col="blue", bg="grey")

## The interface puts data as first parameter
ntbt_roc(aSAH, outcome ~ s100b, type="b", pch=21, col="blue", bg="grey")
ntbt_roc(aSAH, outcome, s100b, type="b", pch=21, col="blue", bg="grey")

## so it can be used easily in a pipeline.
aSAH %>%
  ntbt_roc(outcome ~ s100b, type="b", pch=21, col="blue", bg="grey")
aSAH %>%
  ntbt_roc(outcome, s100b, type="b", pch=21, col="blue", bg="grey")

## End(Not run)

intubate documentation built on May 2, 2019, 2:46 p.m.