compPlot: compPlot: data visualisation comparing observation and model...

In SixiangHu/DataMan: Data Summary, One-way Interactive Plot and Model Visualisation Analysis

Description

This function allows you to visualise different model predictions and observations by certain factors.

Usage

compPlot(x, act, pred, by = NULL, weights = NULL, exposure = NULL, 
   breaks = NULL, missing=TRUE, newGroupNum = 10,xlim=NULL, 
   xname = "x",yname="y",byname="by",legendPos=NULL)

Arguments

x	a vector indicates the dependent variable that you want to visulise on (i.e. Age)
act	a vector indicates the actual response variable (observation)
pred	a vector or data frame that provides model predictions. Each column must be from the same model predictions.
by	Optinal.A numerical vector to specify the <by> variable
weights	Optional. A numerical vector to specify the weights used for calculating weighted average of response. Normally this is the figures from over/down-samping.
exposure	Optional. A numerical vector to specify the exposure used for calculating weighted average of response.
breaks	Optional. A vector to specify the breaks for 'x'
missing	logical. whether to show the 'NA' as 'Missing' in plot. If a 'Missing' level is already existed, then 'NA' will combined
newGroupNum	An integer specifies number of new bands when levels of current plotting variable 'x' or 'by' is more than 100.
xlim	Optional. A vector provides the range of the variable e.g. xlim=c(0,100).
xname, yname, byname	Optional. Characters to be shown on plot.
legendPos	Optional. 2 numeric element vector specify the location of legend.

Author(s)

Sixiang.Hu

Examples

 
compPlot(mtcars$vs,
         act=rnorm(nrow(mtcars)),
         pred = data.frame(pred1=rnorm(nrow(mtcars)),
                           pred2=rnorm(nrow(mtcars))),
         by=mtcars$am)

SixiangHu/DataMan documentation built on May 9, 2019, 1:48 p.m.