Nothing
tests/testthat/test-tTOST.R
In TOSTER: Two One-Sided Tests (TOST) Equivalence Testing
#context("Run Examples for t_TOST")
# need hush function to run print through examples
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
test_that("Run examples for one sample", {
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
set.seed(3164964)
samp1 = rnorm(33)
expect_error(t_TOST())
expect_error(t_TOST(x = samp1,
eqb = c(-1,1,.5)))
expect_error(t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
alpha = 1.22))
expect_error(t_TOST(Sepal.Width ~ Species, data = iris))
# Normal one sample ----
test1 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
alpha = .1)
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
test1 = t_TOST(x = samp1,
eqb = .5)
test1 = t_TOST(x = samp1,
eqb = c(-.5,.5))
test2 = suppressMessages(t_TOST(x = samp1,
#low_eqbound = -.5,
eqb = .5,
eqbound_type = "SMD"))
test3 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( { t_TOST(x = samp1,
#low_eqbound = -.5,
eqb = .5,
eqbound_type = "SMD",
hypothesis = "MET")})
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# Compare to tsum --------
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5)
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5)
tsum1_tg = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5,
glass = "glass1")
tsum1_tg = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5,
glass = "glass2")
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = c(.5,-.5,1)))
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m12 = mean(samp1),
sd12 = sd(samp1),
n12 = length(samp1),
eqb = c(.5,-.5),
paired = TRUE))
expect_warning(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp1),
sd2 = sd(samp1),
r12 = .73,
n2 = 999,
eqb = c(.5,-.5),
paired = TRUE))
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1)))
tsum2 = suppressMessages({ tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
tsum4 = suppressMessages({tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") })
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run -----
test1 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages({ t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE)
})
test3 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( { t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE,
smd_ci = "goulet")
})
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE))
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
prtest = hush(print(test4))
des = hush(describe(test4))
p1 = plot(test4)
p2 = plot(test4,
type = "c")
})
test_that("Run examples for two sample", {
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
tt1 = t_TOST(extra ~ group,
data = sleep,
eqb = 2)
des = describe(tt1)
set.seed(651466441)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
y = samp2,
alpha = .1)
expect_error(smd_calc(x = samp1,
y = samp2,
alpha = -.1))
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
test3 = t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum comparison var.equal = FALSE ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run and non-Welch ----
test1 = t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum comparison var.equal = TRUE ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages({ tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages( {
tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Run with formula
test1 = t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test1_smd = smd_calc(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
bias_correction = FALSE)
# test htest
ash = as_htest(test1)
test2 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
ash = as_htest(test2)
test3 = t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE,
smd_ci = "g") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
prtest = hush(print(test4))
prtest2 = hush(describe(test4))
des = hush(describe(test4))
p1 = plot(test4)
})
test_that("Run examples for paired samples", {
set.seed(789461245)
samp1 = rnorm(25)
samp2 = rnorm(25)
cor12 = stats::cor(samp1,samp2)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
y = samp2,
paired = TRUE,
alpha = .1)
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
ash = as_htest(test1)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
ash = as_htest(test2)
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum paired ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
expect_equal(cor12,
extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1]))
expect_error(extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n = length(samp2),
tstat = NULL))
test_ragain = extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
#sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1])
something = extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
#sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1])
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run and rm_correction
# rm_correction = TRUE
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum paired ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
rm_correction = TRUE,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
rm_correction = TRUE,
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
rm_correction = TRUE,
bias_correction = FALSE)
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
rm_correction = TRUE,
bias_correction = FALSE)})
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Run with formula
test1 = t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
prtest = hush(print(test4))
des = hush(describe(test4))
p1 = plot(test4)
})
test_that("Run examples for plot_smd", {
set.seed(1776)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
p1 = plot_smd(lambda = c(test1$smd$d_lambda),
df = c(test1$smd$d_df),
d = c(test1$smd$d),
type = "cd")
p1 = plot_smd(lambda = c(test1$smd$d_lambda),
df = c(test1$smd$d_df),
d = c(test1$smd$d),
type = "c")
p2 = plot_smd(lambda = c(test2$smd$d_lambda),
df = c(test2$smd$d_df),
d = c(test2$smd$d),
type = "cd")
p2 = plot_smd(lambda = c(test2$smd$d_lambda),
df = c(test2$smd$d_df),
d = c(test2$smd$d),
type = "c")
expect_error(plot_smd(df = c(test1$smd$d_df),
SE = c(test1$smd$d_sigma)))
})
test_that("plot generic function",{
set.seed(1812)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
smd_ci = "g")
expect_error(plot(wilcox_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)))
p1 = plot(test1,
type = "cd",
estimates = "raw")
p2 = plot(test1,
type = "c",
estimates = "raw")
p3 = plot(test1,
type = "cd",
estimates = "SMD")
p4 = plot(test1,
type = "c",
estimates = "SMD")
p5 = suppressMessages(plot(test1,
type = "tnull",
estimates = "SMD"))
p6 = suppressMessages(plot(test1,
type = "tnull"))
p7 = plot(test1,
type = "tnull",
estimates = "raw")
})
test_that("Ensure paired output correct", {
test1 = tsum_TOST(n1 = 23,
n2 = 23,
m2 = 14.2,
m1 = 13.8,
sd1 = 1.23,
sd2 = 1.78,
r12 = .41,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test1$effsize$estimate[1]),sign(test1$effsize$estimate[2]))
test2 = tsum_TOST(n1 = 23,
n2 = 23,
m1 = 14.2,
m2 = 13.8,
sd2 = 1.23,
sd1 = 1.78,
r12 = .41,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test2$effsize$estimate[1]),sign(test2$effsize$estimate[2]))
test3 = t_TOST(extra ~ group, data = sleep,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test3$effsize$estimate[1]),sign(test3$effsize$estimate[2]))
set.seed(90183560)
x1 = rnorm(30)
y1 = rnorm(30)
test4 = t_TOST(x=x1, y=y1,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
test5 = t_TOST(x=x1, y=y1,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = F)
expect_equal(test4$effsize$estimate[2], .0952,
tolerance = .001)
expect_equal(test5$effsize$estimate[2], .0694,
tolerance = .001)
# mean(x1)
# sd(x1)
# mean(y1)
# sd(y1)
# x1: .14 (1.16)
# x2: .04 (1.02)
# r12 = .06
test4 = t_TOST(extra ~ group, data = sleep,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test4$effsize$estimate[1]),sign(test4$effsize$estimate[2]))
})
test_that("Check NCT CIs for paired",{
#effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = TRUE, ci = .9)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = T,
glass = NULL)
expect_equal(test1$effsize$estimate[2],
-1.174, tolerance = .001)
expect_equal(test1$effsize$lower.ci[2],
-1.805, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.4977, tolerance = .001)
#effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = TRUE, ci = .9)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "n",
bias_correction = F,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-1.285, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.975, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.545, tolerance = .001)
test3 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test5_smd = smd_calc(x=subset(sleep, group ==1)$extra,
y=subset(sleep, group ==2)$extra,
paired = TRUE,
smd_ci = "g",
bias_correction = F,
glass = "glass1",
alpha = .1)
test5_smd = smd_calc(x=subset(sleep, group ==1)$extra,
y=subset(sleep, group ==2)$extra,
paired = TRUE,
smd_ci = "g",
bias_correction = F,
glass = "glass2",
alpha = .1)
test6 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
})
test_that("Check NCT CIs for ind",{
#effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd =TRUE)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal= TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-0.7969352, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.523594, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.04942503, tolerance = .001)
#effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal=TRUE,
eqb = .5,
smd_ci = "n",
bias_correction = FALSE,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-0.8321, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.5909, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.05161, tolerance = .001)
test3 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test6 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
## Check avg
#t1=effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd =FALSE)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal= TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-0.7969352, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.523594, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.04942503, tolerance = .001)
#t2=effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd = FALSE)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal=FALSE,
eqb = .5,
smd_ci = "n",
bias_correction = FALSE,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-0.83218, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.59126, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.051069, tolerance = .001)
})
test_that("Check NCT CIs for one",{
z1=subset(sleep, group ==1)$extra-subset(sleep, group ==2)$extra
#t1=effectsize::hedges_g(x=z1,paired = FALSE, ci = .9, pooled_sd =TRUE)
test1 = t_TOST(x=z1,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-1.173925, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.804551, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.4977325, tolerance = .001)
#t2 = effectsize::cohens_d(x=z1,paired = FALSE, ci = .9)
test2= t_TOST(x=z1,
eqb = .5,
smd_ci = "nct",
bias_correction = FALSE)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-1.284558, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.974615, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.5446397, tolerance = .001)
test3 = t_TOST(x=z1,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=z1,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=z1,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test6 = t_TOST(x=z1,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=z1,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=z1,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
})
test_that("More tsum_test",{
expect_error(TOSTER:::tsum_test(
m1 = 12,
sd1 = 1,
n1 = 30,
m2 = 11,
sd2 = 1.5,
n2 = 30,
r12 = NULL,
paired = FALSE,
alternative = "two.sided",
mu = c(0,1),
var.equal = FALSE,
conf.level = 0.95
))
expect_error(TOSTER:::tsum_test(
m1 = 12,
sd1 = 1,
n1 = 30,
m2 = 11,
sd2 = 1.5,
n2 = 30,
r12 = NULL,
paired = FALSE,
alternative = "two.sided",
mu = 0,
var.equal = FALSE,
conf.level = 55
))
})
Try the TOSTER package in your browser
Any scripts or data that you put into this service are public.
TOSTER documentation built on Sept. 15, 2023, 1:09 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#context("Run Examples for t_TOST")
# need hush function to run print through examples
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
test_that("Run examples for one sample", {
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
set.seed(3164964)
samp1 = rnorm(33)
expect_error(t_TOST())
expect_error(t_TOST(x = samp1,
eqb = c(-1,1,.5)))
expect_error(t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
alpha = 1.22))
expect_error(t_TOST(Sepal.Width ~ Species, data = iris))
# Normal one sample ----
test1 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
alpha = .1)
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
test1 = t_TOST(x = samp1,
eqb = .5)
test1 = t_TOST(x = samp1,
eqb = c(-.5,.5))
test2 = suppressMessages(t_TOST(x = samp1,
#low_eqbound = -.5,
eqb = .5,
eqbound_type = "SMD"))
test3 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( { t_TOST(x = samp1,
#low_eqbound = -.5,
eqb = .5,
eqbound_type = "SMD",
hypothesis = "MET")})
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# Compare to tsum --------
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5)
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5)
tsum1_tg = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5,
glass = "glass1")
tsum1_tg = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = .5,
glass = "glass2")
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
eqb = c(.5,-.5,1)))
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m12 = mean(samp1),
sd12 = sd(samp1),
n12 = length(samp1),
eqb = c(.5,-.5),
paired = TRUE))
expect_warning(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp1),
sd2 = sd(samp1),
r12 = .73,
n2 = 999,
eqb = c(.5,-.5),
paired = TRUE))
expect_error(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1)))
tsum2 = suppressMessages({ tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
tsum4 = suppressMessages({tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") })
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run -----
test1 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages({ t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE)
})
test3 = t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( { t_TOST(x = samp1,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE,
smd_ci = "goulet")
})
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages(tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE))
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
prtest = hush(print(test4))
des = hush(describe(test4))
p1 = plot(test4)
p2 = plot(test4,
type = "c")
})
test_that("Run examples for two sample", {
hush = function(code) {
sink("NUL") # use /dev/null in UNIX
tmp = code
sink()
return(tmp)
}
tt1 = t_TOST(extra ~ group,
data = sleep,
eqb = 2)
des = describe(tt1)
set.seed(651466441)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
y = samp2,
alpha = .1)
expect_error(smd_calc(x = samp1,
y = samp2,
alpha = -.1))
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
test3 = t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum comparison var.equal = FALSE ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run and non-Welch ----
test1 = t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum comparison var.equal = TRUE ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages({ tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
tsum4 = suppressMessages( {
tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Run with formula
test1 = t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test1_smd = smd_calc(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
bias_correction = FALSE)
# test htest
ash = as_htest(test1)
test2 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
ash = as_htest(test2)
test3 = t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
var.equal = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE,
smd_ci = "g") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# Check internal consistency
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
prtest = hush(print(test4))
prtest2 = hush(describe(test4))
des = hush(describe(test4))
p1 = plot(test4)
})
test_that("Run examples for paired samples", {
set.seed(789461245)
samp1 = rnorm(25)
samp2 = rnorm(25)
cor12 = stats::cor(samp1,samp2)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)
test1_smd = smd_calc(x = samp1,
y = samp2,
paired = TRUE,
alpha = .1)
expect_equal(test1_smd$estimate, test1$effsize$estimate[2])
expect_equal(test1_smd$lower.ci, test1$effsize$lower.ci[2])
expect_equal(test1_smd$upper.ci, test1$effsize$upper.ci[2])
expect_equal(test1_smd$SE, test1$effsize$SE[2])
ash = as_htest(test1)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
ash = as_htest(test2)
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum paired ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
expect_equal(cor12,
extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1]))
expect_error(extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n = length(samp2),
tstat = NULL))
test_ragain = extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
#sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1])
something = extract_r_paired(m1 = mean(samp1),
sd1 = sd(samp1),
m2 = mean(samp2),
#sd2 = sd(samp2),
n = length(samp2),
tstat = tsum3$TOST$t[1])
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE)})
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Re-run with bias correction not run and rm_correction
# rm_correction = TRUE
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
rm_correction = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
# t_sum paired ----
tsum1 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
rm_correction = TRUE,
bias_correction = FALSE)
tsum2 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
rm_correction = TRUE,
bias_correction = FALSE) })
tsum3 = tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
rm_correction = TRUE,
bias_correction = FALSE)
tsum4 = suppressMessages( { tsum_TOST(m1 = mean(samp1),
sd1 = sd(samp1),
n1 = length(samp1),
m2 = mean(samp2),
sd2 = sd(samp2),
n2 = length(samp2),
r12 = cor12, paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
rm_correction = TRUE,
bias_correction = FALSE)})
expect_equal(test1$TOST$p.value,
tsum1$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test2$TOST$p.value,
tsum2$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test3$TOST$p.value,
tsum3$TOST$p.value,
ignore_attr = TRUE)
expect_equal(test4$TOST$p.value,
tsum4$TOST$p.value,
ignore_attr = TRUE)
# Run with formula
test1 = t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
bias_correction = FALSE)
test2 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
bias_correction = FALSE) )
test3 = t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET",
bias_correction = FALSE)
test4 = suppressMessages( t_TOST(formula = y ~ group,
data = df_samp,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET",
bias_correction = FALSE) )
expect_equal(1-test1$TOST$p.value[2],
test3$TOST$p.value[2])
expect_equal(1-test1$TOST$p.value[3],
test3$TOST$p.value[3])
expect_equal(1-test2$TOST$p.value[2],
test4$TOST$p.value[2])
expect_equal(1-test2$TOST$p.value[3],
test4$TOST$p.value[3])
prtest = hush(print(test4))
des = hush(describe(test4))
p1 = plot(test4)
})
test_that("Run examples for plot_smd", {
set.seed(1776)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
expect_error(t_TOST())
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)
test2 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD") )
test3 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
hypothesis = "MET")
test4 = suppressMessages( t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
eqbound_type = "SMD",
hypothesis = "MET") )
p1 = plot_smd(lambda = c(test1$smd$d_lambda),
df = c(test1$smd$d_df),
d = c(test1$smd$d),
type = "cd")
p1 = plot_smd(lambda = c(test1$smd$d_lambda),
df = c(test1$smd$d_df),
d = c(test1$smd$d),
type = "c")
p2 = plot_smd(lambda = c(test2$smd$d_lambda),
df = c(test2$smd$d_df),
d = c(test2$smd$d),
type = "cd")
p2 = plot_smd(lambda = c(test2$smd$d_lambda),
df = c(test2$smd$d_df),
d = c(test2$smd$d),
type = "c")
expect_error(plot_smd(df = c(test1$smd$d_df),
SE = c(test1$smd$d_sigma)))
})
test_that("plot generic function",{
set.seed(1812)
samp1 = rnorm(25)
samp2 = rnorm(25)
df_samp = data.frame(y = c(samp1,samp2),
group = c(rep("g1",25),
rep("g2",25)))
test1 = t_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5,
smd_ci = "g")
expect_error(plot(wilcox_TOST(x = samp1,
y = samp2,
paired = TRUE,
low_eqbound = -.5,
high_eqbound = .5)))
p1 = plot(test1,
type = "cd",
estimates = "raw")
p2 = plot(test1,
type = "c",
estimates = "raw")
p3 = plot(test1,
type = "cd",
estimates = "SMD")
p4 = plot(test1,
type = "c",
estimates = "SMD")
p5 = suppressMessages(plot(test1,
type = "tnull",
estimates = "SMD"))
p6 = suppressMessages(plot(test1,
type = "tnull"))
p7 = plot(test1,
type = "tnull",
estimates = "raw")
})
test_that("Ensure paired output correct", {
test1 = tsum_TOST(n1 = 23,
n2 = 23,
m2 = 14.2,
m1 = 13.8,
sd1 = 1.23,
sd2 = 1.78,
r12 = .41,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test1$effsize$estimate[1]),sign(test1$effsize$estimate[2]))
test2 = tsum_TOST(n1 = 23,
n2 = 23,
m1 = 14.2,
m2 = 13.8,
sd2 = 1.23,
sd1 = 1.78,
r12 = .41,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test2$effsize$estimate[1]),sign(test2$effsize$estimate[2]))
test3 = t_TOST(extra ~ group, data = sleep,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test3$effsize$estimate[1]),sign(test3$effsize$estimate[2]))
set.seed(90183560)
x1 = rnorm(30)
y1 = rnorm(30)
test4 = t_TOST(x=x1, y=y1,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
test5 = t_TOST(x=x1, y=y1,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = F)
expect_equal(test4$effsize$estimate[2], .0952,
tolerance = .001)
expect_equal(test5$effsize$estimate[2], .0694,
tolerance = .001)
# mean(x1)
# sd(x1)
# mean(y1)
# sd(y1)
# x1: .14 (1.16)
# x2: .04 (1.02)
# r12 = .06
test4 = t_TOST(extra ~ group, data = sleep,
low_eqbound = -.5,
high_eqbound = .5,
paired = T,
bias_correction = FALSE,
eqbound_type = "raw",
rm_correction = T)
expect_equal(sign(test4$effsize$estimate[1]),sign(test4$effsize$estimate[2]))
})
test_that("Check NCT CIs for paired",{
#effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = TRUE, ci = .9)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = T,
glass = NULL)
expect_equal(test1$effsize$estimate[2],
-1.174, tolerance = .001)
expect_equal(test1$effsize$lower.ci[2],
-1.805, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.4977, tolerance = .001)
#effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = TRUE, ci = .9)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "n",
bias_correction = F,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-1.285, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.975, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.545, tolerance = .001)
test3 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test5_smd = smd_calc(x=subset(sleep, group ==1)$extra,
y=subset(sleep, group ==2)$extra,
paired = TRUE,
smd_ci = "g",
bias_correction = F,
glass = "glass1",
alpha = .1)
test5_smd = smd_calc(x=subset(sleep, group ==1)$extra,
y=subset(sleep, group ==2)$extra,
paired = TRUE,
smd_ci = "g",
bias_correction = F,
glass = "glass2",
alpha = .1)
test6 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = TRUE,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
})
test_that("Check NCT CIs for ind",{
#effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd =TRUE)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal= TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-0.7969352, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.523594, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.04942503, tolerance = .001)
#effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal=TRUE,
eqb = .5,
smd_ci = "n",
bias_correction = FALSE,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-0.8321, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.5909, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.05161, tolerance = .001)
test3 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test6 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
## Check avg
#t1=effectsize::hedges_g(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd =FALSE)
test1 = t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal= TRUE,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-0.7969352, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.523594, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.04942503, tolerance = .001)
#t2=effectsize::cohens_d(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,paired = FALSE, ci = .9, pooled_sd = FALSE)
test2= t_TOST(x=subset(sleep, group ==1)$extra,y=subset(sleep, group ==2)$extra,
paired = FALSE,
var.equal=FALSE,
eqb = .5,
smd_ci = "n",
bias_correction = FALSE,
glass = NULL)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-0.83218, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.59126, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.051069, tolerance = .001)
})
test_that("Check NCT CIs for one",{
z1=subset(sleep, group ==1)$extra-subset(sleep, group ==2)$extra
#t1=effectsize::hedges_g(x=z1,paired = FALSE, ci = .9, pooled_sd =TRUE)
test1 = t_TOST(x=z1,
eqb = .5,
smd_ci = "nct",
bias_correction = TRUE)
expect_equal(test1$effsize$estimate[2],
-1.173925, tolerance = .01)
expect_equal(test1$effsize$lower.ci[2],
-1.804551, tolerance = .001)
expect_equal(test1$effsize$upper.ci[2],
-0.4977325, tolerance = .001)
#t2 = effectsize::cohens_d(x=z1,paired = FALSE, ci = .9)
test2= t_TOST(x=z1,
eqb = .5,
smd_ci = "nct",
bias_correction = FALSE)
test_plot = plot(test2)
expect_equal(test2$effsize$estimate[2],
-1.284558, tolerance = .001)
expect_equal(test2$effsize$lower.ci[2],
-1.974615, tolerance = .001)
expect_equal(test2$effsize$upper.ci[2],
-0.5446397, tolerance = .001)
test3 = t_TOST(x=z1,
eqb = .5,
smd_ci = "t",
bias_correction = F,
glass = NULL)
p3 = plot(test3)
test4 = t_TOST(x=z1,
eqb = .5,
smd_ci = "z",
bias_correction = F,
glass = NULL)
p4 = plot(test4)
test5 = t_TOST(x=z1,
eqb = .5,
smd_ci = "g",
bias_correction = F,
glass = "glass1")
test6 = t_TOST(x=z1,
eqb = .5,
smd_ci = "z",
rm_correction = TRUE,
bias_correction = F)
test7 = t_TOST(x=z1,
eqb = .5,
smd_ci = "n",
rm_correction = TRUE,
bias_correction = F)
test8 = t_TOST(x=z1,
eqb = .5,
smd_ci = "t",
rm_correction = TRUE,
bias_correction = F)
})
test_that("More tsum_test",{
expect_error(TOSTER:::tsum_test(
m1 = 12,
sd1 = 1,
n1 = 30,
m2 = 11,
sd2 = 1.5,
n2 = 30,
r12 = NULL,
paired = FALSE,
alternative = "two.sided",
mu = c(0,1),
var.equal = FALSE,
conf.level = 0.95
))
expect_error(TOSTER:::tsum_test(
m1 = 12,
sd1 = 1,
n1 = 30,
m2 = 11,
sd2 = 1.5,
n2 = 30,
r12 = NULL,
paired = FALSE,
alternative = "two.sided",
mu = 0,
var.equal = FALSE,
conf.level = 55
))
})
Try the TOSTER package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.