Manual
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 8
Package
February 17, 2019
Title Calculate Power and Sample Size with BetaPSS
Version 0.0.1
Description Calculate power and sample size(incomplete).
Depends R (>= 3.5.1)
License GPL (>= 2)
Encoding UTF-8
LazyData true
RoxygenNote 6.1.0
Imports betareg,
lmtest,
reshape,
ggplot2,
ggpubr,
Rcpp
Suggests knitr,
rmarkdown
VignetteBuilder knitr
Rtopics documented:
betapower.......................................... 2
betapower_plot ....................................... 3
LFL ............................................. 4
NH.............................................. 5
sample.size2......................................... 5
samplesize_plot....................................... 6
Index 8
1
2betapower
betapower Find Power with Beta DBN
Description
Find the power for a given sample size when testing the null hypothesis that the means for the
control and treatment groups are equal against a two-sided alternative.
Usage
betapower(mu0, sd0, mu1.start, mu1.end, mu1.by,
ss.start, ss.end, ss.by, trials, seed, link.type="logit")
Arguments
mu0 the mean for the control group
sd0 the standard deviation for the control group
mu1.start the starting value of mean for the treatment group under the alternative mu1
mu1.end the ending value of mean for the treatment group under the alternative mu1
mu1.by the step length of mean for the treatment group under the alternative mu1
ss.start the starting value of sample size
ss.end the ending value of sample size
ss.by the step length of sample size
trials the number of trials
seed the seed used in the simulation
link.type the type of link used in the beta regression. Default value is "logit", or you
can choose one or more of the following: "logit", "probit", "cloglog", "cauchit",
"log", "loglog", "all"
Details
betapower function allows you to control the number of trials in the simulation, the sample sizes
used, and the alternative means. You can fix the alternative and vary sample size to match a desired
power; You can fix the sample size and vary the alternative to see which will match a desired power;
You can vary both; Start with a small number of trials (say 100) to determine the rough range of
sample sizes or alternatives; Use a larger number of trials (say 1000) to get better estimates.
Value
Return a matrix with 7 to 12 columns:
power.of.GLM: link name
the power using regression method; it will return the power with every links if
you use link.type = "all" statement.
power.of.Wilcoxon.test
the power from Wilcoxon Rank sum test.
sample size sample size.
mu1 the mean for the treatment group under the alternative.
betapower_plot 3
mu0 the mean for the control group.
sd0 the standard deviation for the control group.
trials the number of trials.
Examples
betapower(0.56,0.255,.70,.75,.05,30,50, 20,40,610201501)
betapower(0.56,0.255,.60,.75,.05,30,50, 5,100,617201501,"all")
betapower(0.56,0.255,.70,.75,.05,30,50, 20,40,610201501,c("logit","loglog","log"))
betapower_plot Plots of Beta power
Description
Generate several comparison plots of power.
Usage
betapower_plot(betapower_matrix,link.type)
Arguments
link.type the type of link used in the beta regression. You can choose one or more of the
following: "logit", "probit", "cloglog", "cauchit", "log", "loglog", "all"
plot.type the type of plot. see details.
betapower_matrix
a matrix obtained by the function betapower.(the formula was described as the
output formula in the function betapower)
Details
betapower_plot() returns different plots depends on plot.type
plot.type = 1: betapower_plot() returns graphs that plot power against mu1, where mu1 is the mean
for the treatment group under the alternative. The number of plots will vary depending on the num-
ber of link types selected with the last plot showing power based on Wilcoxon Rank Sum Test. The
first one or several plots show comparisons of power with different sample size, using GLM method
with one or several link types. The last plot shows a comparison of the power with different sample
size using Wilcoxon Rank Sum Test. Y-axis denotes power and X-axis denotes mu1, the mean for
the treatment group under the alternative.
plot.type = 2: betapower_plot() returns a number of plots equal to the number of sample sizes test-
ed. Each plot compares power calculated with different link types and the Wilcoxon Rank Sum
Test. Y-axis denotes power and X-axis denotes mu1, the mean for the treatment group under the
alternative.
plot.type = 3: betapower_plot() returns a number of plots equal to the number of mu1 used in the
procedure. Each plot compares power calculated with different link types and the Wilcoxon Rank
Sum Test. Y-axis denotes power and X-axis denotes sample size.
4LFL
Examples
BPmat <- betapower(0.56,0.255,.70,.75,.05,30,50, 20,40,610201501,"all")
betapower_plot(BPmat,link.type = "all",plot.type=1)
betapower_plot(BPmat,link.type = "all",plot.type=2)
betapower_plot(BPmat,link.type = "all",plot.type=3)
BPmat2 <- betapower(0.56,0.255,.560,.76,.05,30,45, 5,200,610201511,c("logit","loglog","log"))
betapower_plot(BPmat2,link.type = c("logit","loglog","log"),plot.type=1)
betapower_plot(BPmat2,link.type = c("logit","loglog","log"),plot.type=2)
betapower_plot(BPmat2,link.type = c("logit","loglog","log"),plot.type=3)
LFL Check linearity assumption
Description
Check linearity assumption in a model using a given link function.
Usage
LFL(dsn,Y,X,nkat,link.type)
Arguments
dsn data set name containing X and Y;
Youtcome variable
Xcontinuous covariate
nkat number of categories into which X should be divided (an even number)
link.type link function to be used: link.type is coded as follows: "identity" = identity
link; "logit" = logit link; "probit" = probit link; "log" = log link; "clog-log" =
complementary log-log link; "log-log" = log-log link; "reciprocal" = reciprocal
link;
outlier exclude outliers from the plot.
Value
Return a matrix with 6 columns:
Xintervals of continuous covariate.
Frequency number of observations in each group.
mean group means of outcome variable.
link name of link function.
gmu group means of outcome variable with link function.
midpt median of continuous covariate in each group.
NH 5
Examples
Data.X <- runif(1000)
Data.Y <- exp(Data.X+1)+rnorm(1000,0,1)
Data <- data.frame(cbind(Data.X,Data.Y))
LFL(Data,"Data.Y","Data.X",20,4)
##Use NH data:
NH$quality_rating <- (scale(NH$quality_rating)/2+0.5)
LFL(dsn = NH,Y = "quality_rating", X = "RNHRD", nkat = 40, link.type = "logit", outlier = T)
NH Nursing homes data on Nursing Home Compare site.
Description
These are the official datasets used on the Medicare.gov Nursing Home Compare Website provided
by the Centers for Medicare & Medicaid Services. These data allow you to compare the quality of
care at every Medicare and Medicaid-certified nursing home in the country, including over 15,000
nationwide. This data set only used 3 variables of original data.
Usage
data(NH)
Format
An object of data frame.
Source
Nursing homes data
Examples
data(NH)
summary(NH)
sample.size2 Find minimum sample size with Beta DBN
Description
Find minimum sample sizes with Beta DBN and given mu0,sd0,mu1 and target powers.
Usage
sample.size(mu0, sd0, mu1.start, mu1.end, mu1.by, power.start, power.end, power.by,
sig.level = 0.05, N = 100, accuracy = NuLL, seed = 1, link.type = "logit")
6samplesize_plot
Arguments
mu0 the mean for the control group
sd0 the standard deviation for the control group
mu1.start the starting value of mean for the treatment group under the alternative mu1
mu1.end the ending value of mean for the treatment group under the alternative mu1
mu1.by the step length of mean for the treatment group under the alternative mu1
power.start the starting value of target power
power.end the ending value of target power
power.by the step length of target power
sig.level significant level; default value is 0.05
Nthe number of trials; default value is 100
accuracy the accuracy of the result; must be integer
seed the seed used in the simulation
link.type link options include: "logit", "probit", "cloglog", "cauchit", "log", "loglog". De-
fault link is "logit".
Details
The sample.size function allows you to control the number of trials in the simulation, the target
power, accuracy, and the alternative means. You can fix the alternative and vary power to match a
desired sample size; Use default values for the number of trials and accuracy for a quick view; Use
a larger number of trials (say 1000) and a smaller accuracy (say 1) to get better estimates.
Examples
sample.size2(mu0=0.56, sd0=0.255, mu1.start = 0.60, mu1.end = 0.70, mu1.by = 0.05,
power.start = 0.7, power.end = 0.9, power.by = 0.1)
sample.size2(mu0=0.56, sd0=0.255, mu1.start = 0.60, mu1.end = 0.70, mu1.by = 0.05,
power.start = 0.7, power.end = 0.9, power.by = 0.1, link.type = c("logit","loglog","log"))
sample.size2(mu0=0.56, sd0=0.255, mu1.start = 0.60, mu1.end = 0.70, mu1.by = 0.05,
power.start = 0.7, power.end = 0.9, power.by = 0.1, link.type = "all")
samplesize_plot Plots by mu1
Description
Generate the comparison plots using GLM method and Wilcoxon Rank Sum Test with different
mu1.
Usage
samplesize_plot(SS_matrix,link.type)
samplesize_plot 7
Arguments
link.type the type of link used in the beta regression(or Wilcoxon Rank Sum Test). You
can choose one or more of the following: "logit", "probit", "cloglog", "cauchit",
"log", "loglog", "wilcoxon", "all"
SS_matrix the matrix obtained by the function sample.size2.(the formula was described as
the output formula in the function sample.size2)
Details
samplesize_plot() returns a series of plots equal to the number of mu1 used in the procedure. Y-axis
denotes minimum sample size and X-axis denotes minimum power.
Examples
SSmat <- sample.size2(mu0=0.56, sd0=0.255, mu1.start = 0.60, mu1.end = 0.70, mu1.by = 0.05,
power.start = 0.7, power.end = 0.9, power.by = 0.1, link.type = "all")
samplesize_plot(SSmat, "all")
SSmat2 <- sample.size2(mu0=0.56, sd0=0.255, mu1.start = 0.60, mu1.end = 0.70, mu1.by = 0.05,
power.start = 0.7, power.end = 0.9, power.by = 0.1, link.type = c("logit","loglog","log"))
samplesize_plot(SSmat2,link.type = c("logit","loglog","log"))
