sprtt

# installs the package
install.packages("sprtt")

# the installation requires the "devtools" package
# install.packages("devtools")
# stable GitHub version
devtools::install_github("MeikeSteinhilber/sprtt")

# development version, may not be stable
devtools::install_github("MeikeSteinhilber/sprtt", ref="develop")

# set seed --------------------------------------------------------------------
set.seed(333)

# load library ----------------------------------------------------------------
library(sprtt)
#> Note: Sample size planning functions require simulation data (~150 MB).
#> Data will download automatically on first use.
#> See ?download_sample_size_data for more information.

# t-TEST ----------------------------------------------------------------------
# one sample: numeric input ---------------------------------------------------
treatment_group <- rnorm(20, mean = 0, sd = 1)
results <- seq_ttest(treatment_group, mu = 1, d = 0.6)

# get access to the slots -----------------------------------------------------
# @ Operator
results@likelihood_ratio
#> [1] 323.9051

# [] Operator
results["likelihood_ratio"]
#> [1] 323.9051

# ANOVA -----------------------------------------------------------------------
# simulate data ---------------------------------------------------------------
set.seed(333)
data <- sprtt::draw_sample_normal(k_groups = 3,
                                  f = 0.15,
                                  sd = c(1, 1, 1),
                                  max_n = 46)
# look at the data
head(data)
#>            y x
#> 1  0.2762330 1
#> 2 -1.3415005 2
#> 3 -0.4521177 3
#> 4  1.2315479 1
#> 5  0.8180678 2
#> 6  0.1697634 3

# calculate sequential ANOVA --------------------------------------------------
results <- sprtt::seq_anova(y ~ x, f = 0.15, data = data, plot = TRUE)
# test decision
results@decision
#> [1] "accept H1"
# test results
results
#> 
#> *****  Sequential ANOVA *****
#> 
#> formula: y ~ x
#> test statistic:
#>  log-likelihood ratio = 3.061, decision = accept H1
#> SPRT thresholds:
#>  lower log(B) = -2.944, upper log(A) = 2.944
#> Log-Likelihood of the:
#>  alternative hypothesis = -3.839
#>  null hypothesis = -6.9
#> alternative hypothesis: true difference in means is not equal to 0.
#> specified effect size: Cohen's f = 0.15
#> empirical Cohen's f = 0.3255325, 95% CI[0.1324216, 0.4835088]
#> Cohen's f adjusted = 0.296
#> degrees of freedom: df1 = 2, df2 = 135
#> SS effect = 12.83987, SS residual = 121.1635, SS total = 134.0034
#> *Note: to get access to the object of the results use the @ or [] instead of the $ operator.

# plot results -----------------------------------------------------------------
sprtt::plot_anova(results)

# save the plot ----------------------------------------------------------------
ggplot2::ggsave(
  "man/figures/readme_example.png",       # define the path
   units = "cm", height = 10, width = 16, # define the image size
   dpi = 300                              # define resolution
  )

sprtt::plan_sample_size(f_expected = 0.25,
                        k_groups = 3,
                        beta = 0.05,
                        decision_rate = 0.85)

# Check cache status
sprtt::cache_info()

# Force re-download (if data was updated)
sprtt::download_sample_size_data(force = TRUE)

# Clear cache
sprtt::cache_clear()