The sprtt package is the implementation of sequential probability ratio tests using the associated t-statistic (sprtt).
The package contains:
seq_ttest() calculates the sequential test statistic and
three data sets (df_income, df_stress, df_cancer) to run the examples in the documentation
This is the recommended version for a normal user.
To get a bug fix or to use a feature from the development version, you can install the development version from GitHub.
# the installation requires the "devtools" package
# install.packages("devtools")
devtools::install_github("MeikeSteinhilber/sprtt")Detailed documentation can be found on the home page. There are several articles covering the usage of the package, the theoretical background of the test, and also an extended use case.
Short examples can be found in the following paragraph.
Note
In the R code sections:
# comment: is a comment
function(): is R code
#> results of function(): is console output
# set seed --------------------------------------------------------------------
set.seed(333)
# load library ----------------------------------------------------------------
library(sprtt)
# one sample: numeric input ---------------------------------------------------
treatment_group <- rnorm(20, mean = 0, sd = 1)
results <- seq_ttest(treatment_group, mu = 1, d = 0.8)
# get access to the slots -----------------------------------------------------
# @ Operator
results@likelihood_ratio
#> [1] 965.0728
# [] Operator
results["likelihood_ratio"]
#> [1] 965.0728
# two sample: numeric input----------------------------------------------------
treatment_group <- stats::rnorm(20, mean = 0, sd = 1)
control_group <- stats::rnorm(20, mean = 1, sd = 1)
seq_ttest(treatment_group, control_group, d = 0.8)
#>
#> ***** Sequential Two Sample t-test *****
#>
#> data: treatment_group and control_group
#> test statistic:
#> log-likelihood ratio = 5.347, decision = accept H1
#> SPRT thresholds:
#> lower log(B) = -2.94444, upper log(A) = 2.94444
#> Log-Likelihood of the:
#> alternative hypothesis = -4.21063
#> null hypothesis = -9.55763
#> alternative hypothesis: true difference in means is not equal to 0.
#> specified effect size: Cohen's d = 0.8
#> degrees of freedom: df = 38
#> sample estimates:
#> mean of x mean of y
#> -0.05204 1.18768
#> Note: to get access to the object of the results use the @ or []
#> instead of the $ operator.
# two sample: formula input ---------------------------------------------------
stress_level <- stats::rnorm(20, mean = 0, sd = 1)
sex <- as.factor(c(rep(1, 10), rep(2, 10)))
seq_ttest(stress_level ~ sex, d = 0.8)
#>
#> ***** Sequential Two Sample t-test *****
#>
#> data: stress_level ~ sex
#> test statistic:
#> log-likelihood ratio = -1.45506, decision = continue sampling
#> SPRT thresholds:
#> lower log(B) = -2.94444, upper log(A) = 2.94444
#> Log-Likelihood of the:
#> alternative hypothesis = -1.23287
#> null hypothesis = 0.2222
#> alternative hypothesis: true difference in means is not equal to 0.
#> specified effect size: Cohen's d = 0.8
#> degrees of freedom: df = 18
#> sample estimates:
#> mean of x mean of y
#> -0.23286 -0.08217
#> Note: to get access to the object of the results use the @ or []
#> instead of the $ operator.
# NA in the data --------------------------------------------------------------
stress_level <- c(NA, stats::rnorm(20, mean = 0, sd = 2), NA)
sex <- as.factor(c(rep(1, 11), rep(2, 11)))
seq_ttest(stress_level ~ sex, d = 0.8, na.rm = TRUE)
#>
#> ***** Sequential Two Sample t-test *****
#>
#> data: stress_level ~ sex
#> test statistic:
#> log-likelihood ratio = -0.3585, decision = continue sampling
#> SPRT thresholds:
#> lower log(B) = -2.94444, upper log(A) = 2.94444
#> Log-Likelihood of the:
#> alternative hypothesis = -1.923
#> null hypothesis = -1.5645
#> alternative hypothesis: true difference in means is not equal to 0.
#> specified effect size: Cohen's d = 0.8
#> degrees of freedom: df = 18
#> sample estimates:
#> mean of x mean of y
#> -0.40818 0.42068
#> Note: to get access to the object of the results use the @ or []
#> instead of the $ operator.
# work with dataset (data are in the package included) ------------------------
seq_ttest(monthly_income ~ sex, data = df_income, d = 0.8)
#>
#> ***** Sequential Two Sample t-test *****
#>
#> data: monthly_income ~ sex
#> test statistic:
#> log-likelihood ratio = -9.51391, decision = accept H0
#> SPRT thresholds:
#> lower log(B) = -2.94444, upper log(A) = 2.94444
#> Log-Likelihood of the:
#> alternative hypothesis = -8.09254
#> null hypothesis = 1.42137
#> alternative hypothesis: true difference in means is not equal to 0.
#> specified effect size: Cohen's d = 0.8
#> degrees of freedom: df = 118
#> sample estimates:
#> mean of x mean of y
#> 3072.086 3080.715
#> Note: to get access to the object of the results use the @ or []
#> instead of the $ operator.