plot.UPSnnltd {USPS}R Documentation

Display plots of the NN/LTD Distribution in Unsupervised Propensiy Scoring

Description

Make a Snow Ball ("snob"), Probability Density ("dens") and/or Cumulative Distribution Function ("cdf") plot of the Local Treatment Difference (LTD) Distribution across all Informative Clusters within a requested total number clusters in patient baseline X-covariate space.

Usage

  ## S3 method for class 'UPSnnltd':
  plot(x, pballs = TRUE, nnplot = "snob", nnalpha = 1.4, ...)

Arguments

x Object of class UPSnnltd.
pballs Logical; Display cluster-size weighted (red) and inverse-variance weighted (green) "Power Balls" to represent across cluster LTD averages within the "Snowball" plot.
nnplot String; Display "all" plots together, "seq" = one after the other, "snob" = only Snow Balls, "dens" = only fitted gss probability density, or "cdf" = only the gss cumulative distribution.
nnalpha Numeric; alpha argument to ssden() function.
... Optional argument(s) to pass on to plot().

Details

Plots of UPSnnltd objects describe the LTD Distribution across all informative clusters. The generalized smoothing spline (gss) displays corresponding to the nnplot = "dens" and "cdf" options may fail to exist when the number of informative clusters is less than 60.

Value

NULL

Author(s)

Bob Obenchain <wizbob@att.net>

References

Obenchain RL. (2004) Unsupervised Propensity Scoring: NN and IV Plots. Proceedings of the American Statistical Association (on CD) 8 pages.

Obenchain RL. (2009) USPSinR.pdf ../R_HOME/library/USPS 40 pages.

See Also

UPSivadj, UPSaccum and UPSgraph.

Examples

    data(lindner)
    UPSxvars <- c("stent", "height", "female", "diabetic", "acutemi", "ejecfrac", "ves1proc")
    UPSharch <- UPShclus(lindner, UPSxvars)
    UPSaccum(UPSharch, lindner, abcix, lifepres, faclev=1, scedas="homo", accobj="ABClife")
    
    lif070nn <- UPSnnltd(70)
    lif070nn
    plot(lif070nn)

[Package USPS version 1.2-0 Index]