inRange

Description

Return which values are in a certain range %inRange% indicates which values are in a certain range, including the boundaries of the range.

Usage

lhs %inRange% rhs

Arguments

Value

%inRange% returns a logical vector of length(lhs), indicating which values of lhs are and are not in range rhs. Boundaries of rhs are included.

Author(s)

Mathijs Deen

See Also

%withinRange%

Examples

a <- seq(0, 100, 5)
r <- c(40, 70)
cbind(a,
      'a %inRange% r' = a %inRange% r,
      'a %withinRange% r' = a %withinRange% r)

Inverse value matching

Description

Evaluates whether the left hand side argument is not in the right hand side argument.

Usage

lhs %ni% rhs

Arguments

Details

The %ni% function negates the %in% operator.

Value

The function returns a vector of the same length as lhs.

Author(s)

Mathijs Deen

Examples

c(1,2,3) %ni% c(1,2)

withinRange

Description

Return which values are within a certain range

%withinRange% indicates which values are in a certain range, excluding the boundaries of the range.

Usage

lhs %withinRange% rhs

Arguments

Value

%withinRange% returns a logical vector of length(lhs), indicating which values of lhs are and are not in range rhs. Boundaries of rhs are excluded.

Author(s)

Mathijs Deen

See Also

%inRange%

Examples

a <- seq(0, 100, 5)
r <- c(40, 70)
cbind(a,
      'a %inRange% r' = a %inRange% r,
      'a %withinRange% r' = a %withinRange% r)

Cost-effectiveness analysis

Description

Perform a cost-effectiveness analysis. Or a cost-utility analysis.

Usage

CEA(data, group, cost, effect, B = 5000, currency = "euro")

Arguments

Value

CEA returns a list (class CEA) with the following elements:

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 5000, "acorns") |>
  plot()

Cost-effectiveness acceptability curve

Description

Create data for cost-effectiveness acceptability curve

Usage

CEAC(x)

Arguments

Value

CEAC returns data that can be plotted using plot.CEAC.

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  plot(xlim = c(0,200))

QIDS depression data

Description

The QIDS dataset consists of 100 observations of people that were clinically diagnosed with a major depressive disorder and who filled out the QIDS-SR questionnaire. The data were simulated.

Usage

QIDS

Format

the following variables are available:

• QIDS: QIDS-SR total score.

• Depression: an indicator whether the individual was diagnosed with a depression or not.

Author(s)

Mathijs Deen

Calculate (pseudo) R^2 for vglm objects

Description

Calculate (pseudo) R^2 for vglm objects

Usage

R2.vglm(
  model,
  method = c("mcfadden", "nagelkerke", "efron", "coxsnell", "tjur")
)

Arguments

Value

R2.vglm returns R^2.

Author(s)

Mathijs Deen

Examples

if(require("VGAM")){
  fit <- vglm(Species ~ Sepal.Length,
              family = multinomial(),
              data = iris)
  R2.vglm(fit)
}

Area under the curve

Description

Calculate the area under the curve.

Usage

auc(x, ...)

Arguments

Value

returns the area under the curve for a roc class object.

Author(s)

Mathijs Deen

Examples

a <- roc(QIDS$QIDS, QIDS$depression, c("Yes","No"), "Yes")
auc(a)

Checks for linear model

Description

Perform checks for a linear model regarding influential cases and collinearity numerically and graphically.

Usage

check(object, ...)

Arguments

Value

check returns a list containing two matrices with statistics regarding influential cases and a vector of variance inflation factors. Furthermore, it produces diagnostics plots.
The return list contains three elements:

- influence, a data.frame, with observations in the model, and the following variables:

- is.infl is a data.frame indicating which influence measure(s) is/are flagged per observation.

- vifs, a vector containing variance inflation factors for the variables in the model.

By default, the two data.frames regarding influence measures only give the influence measures for cases that are flagged as being influential. Influence measures for all cases can be queried using print.check.lm.

The generated plots are the plots produced by plot.lm, numbers 1 through 6.

influential cases

For the influence indicators, the following rules are applied to check whether a case is influential:

• \mathrm{any\enspace}|\mathrm{dfbeta}| > 1.

• |\mathrm{dffit}| > 3 \sqrt{\frac{k}{n-k}}.

• |1 - \mathrm{cov.r}| > \frac{3k}{n-k}.

• F\mathrm{(}n, n-k \mathrm{)} = \mathrm{cooks.d\enspace having\enspace}. p > .5

• \mathrm{hat} > \frac{3k}{n}.

These indicators for being an influential case were derived from influence.measures in the stats package.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ disp + wt, data = mtcars)
check(lm.1)

Coefficients for logistic regression analysis

Description

Show odds ratios and their confidence intervals for logistic regression parameter estimates.

Usage

coefsLogReg(model, confint = TRUE, level = 0.95)

Arguments

Value

coefsLogReg returns the same table as summary(object)$coefficients, with the addition of the coefficients' odds ratios and their confidence intervals.

Author(s)

Mathijs Deen

Examples

glm(formula = am ~  disp, family = binomial, data = mtcars) |>
 coefsLogReg()

List of correlation coefficients

Description

List all correlations in a correlation matrix without duplicates.

Usage

corList(x, ...)

Arguments

Value

corList returns a list of correlations

Author(s)

Mathijs Deen

Examples

mtcars[,c("mpg","disp", "hp", "drat", "wt", "qsec")] |>
  corList(method="spearman")

Currency unicode character

Description

Retrieve Unicode character for a currency

Usage

currency2unicode(currency, type = c("character", "code"))

Arguments

Details

The input is evaluated case insensitive. In case the input is not supported, the function will return the original input.

Value

currency2unicode the unicode character for a given currency.

Author(s)

Mathijs Deen

Examples

currency2unicode("dollar")

cat(sprintf("%s5 is all my mom allows me to spend.", currency2unicode("dollar")))

Effect sizes from pretest-posttest-control group designs

Description

dPPC2 calculates an effect size for studies with pretest and posttest scores for two groups, usually a treatment and a control group. It is based on Morris (2008), who based it on Becker (1988).

Usage

dPPC2(preT, posT, preC, posC, correct = TRUE, CIlevel = 0.95)

Arguments

Value

dPPC2 returns a vector of length 6, containing:

Author(s)

Mathijs Deen

References

• Becker, B.J. (1988). Synthesizing standardized mean-change measures. British Journal of Mathematical and Statistical Psychology, 41, 257-278.

• Morris, S.B. (2008). Estimating effect sizes from pretest-posttest-control group designs. Organizational Research Methods, 11, 364-386.

Examples

library(MASS)
set.seed(1)
treatment <- mvrnorm(n=50, mu=c(50,40), Sigma = matrix(c(100,70,70,100), ncol=2), empirical = TRUE)
control <- mvrnorm(n=50, mu=c(50,45), Sigma = matrix(c(100,70,70,100), ncol=2), empirical = TRUE)
dPPC2(treatment[,1], treatment[,2], control[,1], control[,2])

Local f^2

Description

Calculate local f^2 for (generalized) linear (mixed) models

Usage

f2Local(object, method, ...)

## S3 method for class 'lm'
f2Local(object, method = "r.squared", ...)

## S3 method for class 'glm'
f2Local(object, method = "r2", ...)

## S3 method for class 'vglm'
f2Local(object, method = "mcfadden", ...)

## S3 method for class 'glmmTMB'
f2Local(object, method = "nakagawa", type = "marginal", ...)

Arguments

Details

The following methods can be specified:

• lm objects: r.squared and adj.r.squared as extracted from the lm object.

• glm objects: mcfadden, nagelkerke, coxsnell, tjur and efron, as implemented in the performance package.

• vglm objects: mcfadden, nagelkerke, coxsnell, tjur and efron, as implemented in the R2.vglm function.

• glmmTMB objects: nakagawa, as implemented in the performance package. It can also be specified whether the marginal or the conditional R^2 should be used, however only the marginal R^2 would make sense.

Note that for multinomial models, using method="efron" gives questionable with glm objects and is not possible for vglm objects. For glm objects, method=coxsnell cannot be used when the response is not binary.

Value

f2Local returns a list containing f^2 values for every parameter in a model. For the glmmTMB class, a list of all reduced models is returned as well. In a future version, this will be available for other classes as well.

Methods (by class)

• f2Local(lm): Method for lm object

• f2Local(glm): Method for glm object

• f2Local(vglm): Method for vglm object

• f2Local(glmmTMB): Method for glmmTMB object

Author(s)

Mathijs Deen

Examples

# linear model
model1 <- lm(mpg ~ cyl + wt*drat, data = mtcars)
f2Local(model1)

# generalized linear model (glm)
model2 <- glm(vs ~  cyl*wt + mpg, data = mtcars, family = "binomial")
f2Local(model2, method = "coxsnell")

# generalized linear model (vglm)
if(require(VGAM)){
  pneumo <- transform(pneumo, let = log(exposure.time))
  model3 <- vglm(cbind(normal, mild, severe) ~ let, multinomial, pneumo)
  f2Local(model3)
}
# generalized linear mixed model
if(require(ClusterBootstrap) & require(glmmTMB)){
  model4 <- glmmTMB(pos ~ treat*time + (1 + time | id), data = medication)
  f2Local(model4)
}

Display frequency table

Description

Display frequency table with percentages and cumulative percentages.

Usage

frequencies(x)

Arguments

Value

object of type data.frame containing frequencies, percentages and cumulative percentages.

Author(s)

Mathijs Deen

Examples

frequencies(datasets::mtcars$carb)

Gnomes mushroom insulation cost-effectiveness data

Description

The gnomes dataset consists of 300 observations of gnomes that had their housing unit (i.e., their mushroom) insulated against cold and humidity. The insulation was done by the most skilled insulation animals in the forest: squirrels (Sciurus vulgaris).They either used the common insulation technique consisting of leafs of the common beech tree (Fagus sylvatica) or an experimental form of insulation using leafs of the less common (and thus, more expensive) sessile oak tree (Quercus petraea). For the year before insulation and the year after insulation, the gnomes filled out the Gnomes' Humidity and Thermal Satisfaction scale (Gnomes' HATS), a well-validated questionnaire that rates mushroom insulation satisfaction w.r.t. humidity and temperature on a scale of 0 to 50. Differences between pre and post measurement were calculated on a higher-is-better basis.

The squirrels were paid in acorns.

Usage

gnomes

Format

the following variables are available:

• diffHATS: the difference in Gnomes' HATS scores between the year before and the year after insulation.

• Costs: insulation costs in acorns.

• insulationMethod: method of insulation, either commonBeech or sessileOak.

Author(s)

Mathijs Deen

Save something to an object

Description

keep saves an object to a new object. This is useful if one wants to save an intermediate result when using pipes.

Usage

keep(object, name, pos = 1, envir = as.environment(pos), inherits = FALSE)

Arguments

Value

Upon saving object to name, the value of object is returned. This makes it suitable for pipes.

Author(s)

Mathijs Deen

Examples

mtcars |>
  lm(mpg ~  disp + hp, data = _) |>
  keep(lm.mpg_disp_hp) |>
  summary()

Mean center

Description

Mean center a vector or numeric matrix.

Usage

m(x)

Arguments

Details

This function resembles base::scale.default, with the scale argument set to FALSE. This, together with the short function name, is especially useful when you want to mean center variables in an analysis (e.g., using (g)lm), but you dont want the long form scale(x, scale=FALSE) to clutter up the rownames of the parameter estimates or the model anova.

Value

m returns a mean centered version of x. If x is a matrix, the matrix dimensions are preserved.

Author(s)

Mathijs Deen

Examples

vals <- matrix(rnorm(24, 15, 10), ncol = 2)
m(vals)

Posterior model probability

Description

Calculate the posterior model probability for a set of models.

Usage

pMM(...)

Arguments

Details

Posterior model probabilities are calculated for every model i as

\mathrm{pMO}_i = \frac{\mathrm{exp}\Big[-\frac{1}{2}\Delta_i\mathrm{BIC}\Big]}{\sum_{j = 1}^K\mathrm{exp}\Big[-\frac{1}{2}\Delta_j\mathrm{BIC}\Big]},

where the minimal BIC value is subtracted from all BICs. In other words: the model with the lowest BIC has \Delta\mathrm{BIC}=0.

Value

pMM returns to posterior model probabilities for the models provided.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ hp + wt, data = mtcars)
lm.2 <- lm(mpg ~ hp * wt, data = mtcars)
lm.3 <- lm(mpg ~ hp * wt + gear, data = mtcars)
pMM(lm.1, lm.2, lm.3)

Plot cost-effectiveness plane

Description

Plot cost-effectiveness plane

Usage

## S3 method for class 'CEA'
plot(
  x,
  xlim = c(-1, 1) * max(abs(x$stats$diffE)),
  ylim = c(-1, 1) * max(abs(x$stats$diffC)),
  xlab = "Incremental effects",
  ylab = sprintf("Incremental costs (%s)", x$currencyUC),
  las = 1,
  ...
)

Arguments

Value

plot.CEA returns a plot

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 5000, "acorns") |>
  plot()

Plot cost-effectiveness acceptability curve

Description

Plot cost-effectiveness acceptability curve

Usage

## S3 method for class 'CEAC'
plot(
  x,
  xlab = sprintf("Cost-effectiveness threshold (%s)", x$currencyUC),
  ylab = "Probability that intervention is cost-effective",
  las = 1,
  xlim = c(0, max(x$s$ICERs)),
  ...
)

Arguments

Value

returns a plot

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  plot(xlim = c(0,200))

plot probed interaction

Description

Plot the effects of the antecedent as a function of the moderator.

Usage

## S3 method for class 'probeInteraction'
plot(
  x,
  ...,
  col.JN = "red",
  lty.JN = "dotted",
  col.CI = rgb(red = 0.5, green = 0.5, blue = 0.5, alpha = 0.2),
  lty.CI = "longdash",
  lty.0 = "dotted"
)

Arguments

Value

plot.probeInteraction returns a combined plot with p value on the first y axis and effect of the antecedent variable.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ hp * wt, data = mtcars)
pI.1 <- probeInteraction(lm.1, hp, wt, JN=TRUE, n.interval.moderator = 3,
                         quantile.moderator = c(0.1,0.9), values.moderator = 2)
plot(pI.1)
lm.2 <- lm(mpg ~ qsec * drat, data = mtcars)
pI.2 <- probeInteraction(lm.2, qsec, drat, JN=TRUE, n.interval.moderator = 30,
                         quantile.moderator = c(0.1,0.9), values.moderator = 2)
plot(pI.2)

plot roc curve

Description

Plot an ROC curve.

Usage

## S3 method for class 'roc'
plot(
  x,
  y,
  which = 1:3,
  orientation = c("horizontal", "vertical"),
  cutoffs.1 = NULL,
  cutoffs.2 = NULL,
  cutoffs.3 = NULL,
  xlab.3 = NULL,
  labels.3 = NULL,
  xlim.3 = NULL,
  ylim.3 = c(0, 10),
  pos.legend.2 = "right",
  pos.legend.3 = "topright",
  ...
)

Arguments

Value

plot.roc provides three plots:

• The first plot contains the ROC curve.

• The second plot contains curves for the sensitivity and the specificity for all threshold values.

• The third plot contains density plots for the two classification groups.

Author(s)

Mathijs Deen

Examples

a <- roc(QIDS$QIDS, QIDS$depression, c("Yes","No"), "Yes")
plot(a, ylim.3 = c(0,.2), xlab.3= "QIDS value", cutoffs.1 = 14.5,
     cutoffs.2 = 14.5, cutoffs.3 = 14.5)

Plot a probability distribution

Description

Plot the density function of certain probability distributions.

Usage

plotDistribution(
  distribution = c("normal", "t", "chi2", "F"),
  xRange = c(0, 5),
  xColArea = NULL,
  xAreaCol = NULL,
  mean = 0,
  sd = 1,
  df,
  df1,
  df2,
  ncp,
  ...
)

Arguments

Value

plotDistribution returns a probability density plot.

Author(s)

Mathijs Deen

Examples

plotDistribution(distribution = "normal",
                 xRange       = c(-5, 5),
                 xColArea     = matrix(data  = c(-5, -1.96,
                                                 1.96, 5),
                                       ncol  = 2,
                                       byrow = TRUE),
                 xAreaCol     = c("green", "blue"),
                 mean         = 0,
                 sd           = 1,
                 yaxt         = "n",
                 ylab         = "")

Print lm check

Description

Print the check of lm object

Usage

## S3 method for class 'check.lm'
print(x, which.infl = c("influential", "all"), ...)

Arguments

Value

prints the check.lm object.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ disp + wt, data = mtcars)
chk.lm.1 <- check(lm.1)
print(chk.lm.1, which.infl="all")

Print effects of probed interaction

Description

Print the effects from a probed interaction.

Usage

## S3 method for class 'probeInteraction'
print(x, ...)

Arguments

Value

print.probeInteraction prints the effects table of a probeInteraction object.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ hp * wt, data = mtcars)
pI <- probeInteraction(lm.1, hp, wt, JN=TRUE, n.interval.moderator = 3,
                       quantile.moderator = c(0.1,0.9), values.moderator = 2)
print(pI)

Print t test

Description

Print the output of a t test.

Usage

## S3 method for class 'tTest'
print(x, ...)

Arguments

Value

prints the tTest object as a htest object.

Author(s)

Mathijs Deen

Examples

x1 <- QIDS$QIDS[QIDS$depression == "Yes"]
x2 <- QIDS$QIDS[QIDS$depression == "No"]
tt <- tTest(x1, x2)
print(tt)

Print willingness to pay probe

Description

Print the outcome of a willingness to pay threshold probe.

Usage

## S3 method for class 'wtp'
print(x, ...)

Arguments

Value

print.wtp prints the outcome of wtp

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  wtp(probability = 0.80)
CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  wtp(threshold = 8)

Probe interaction models

Description

Probe the effect of a moderator on an X/antecedent variable in a linear model.

Usage

probeInteraction(
  object,
  antecedent,
  moderator,
  alpha = 0.05,
  JN = TRUE,
  n.interval.moderator,
  quantile.moderator,
  values.moderator
)

Arguments

Details

the arguments n.interval.moderator, quantile.moderator and values.moderator can be combined. All unique values from these methods combined, together with the values from the Johnson-Neyman procedure (if specified) will be part of the probing procedure.

Value

probeInteraction returns a data frame containing values of the moderator in a linear model, the effect of the antecedent at that value of the moderator, standard errors, t values, p values and a confidence interval.

Author(s)

Mathijs Deen

Examples

lm.1 <- lm(mpg ~ hp * wt, data = mtcars)
probeInteraction(lm.1, hp, wt, JN=TRUE, n.interval.moderator = 3,
                 quantile.moderator = c(0.1,0.9), values.moderator = 2)

Reliable change index

Description

rci computes the reliable change index according to Jacobson and Truax (1992).

Usage

rci(x1, x2, rxx)

Arguments

Value

rci returns a vector of length(x1) with reliable change index scores.

Author(s)

Mathijs Deen

References

• Jacobson, N.S., & Truax, P. (1992). Clinical significance: a statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12-19.

Examples

library(MASS)
set.seed(1)
q <- mvrnorm(n=120, mu=c(40, 50), Sigma = matrix(c(56.25,45,45,56.25), ncol = 2), empirical = TRUE)
cbind(q, rci(q[,1], q[,2], .8), rci(q[,1], q[,2], .8) > 1.96)

Receiver operator characteristic

Description

Calculate ROC curve statistics.

Usage

roc(response, group, levels, state)

Arguments

Value

Returns a list with the following elements:

Author(s)

Mathijs Deen

Examples

roc(QIDS$QIDS, QIDS$depression, c("No","Yes"), "Yes") |>
  plot(ylim.3=c(0,.2))

Summarize outcome of a t test

Description

Summarize the outcome of a t test

Usage

## S3 method for class 'tTest'
summary(object, rnd = 3L, ...)

Arguments

Value

summary.htest returns a typical APA-like output (without italics) for a t-test.

Author(s)

Mathijs Deen

Examples

x1 <- QIDS$QIDS[QIDS$depression == "Yes"]
x2 <- QIDS$QIDS[QIDS$depression == "No"]
tt <- tTest(x1, x2)
summary(tt, rnd = c(1,2,3))

t Test

Description

perform t tests with the possibility of inputting group statistics.

Usage

tTest(
  x,
  y = NULL,
  sdx = NULL,
  sdy = NULL,
  nx = length(na.omit(x)),
  ny = length(na.omit(y)),
  alternative = c("two.sided", "greater", "less"),
  mu = 0,
  paired = FALSE,
  rxy = NULL,
  var.equal = FALSE,
  conf.level = 0.95
)

Arguments

Value

tTest performs a t-test (independent samples, paired samples, one sample) just like base-R t.test, but with the extended possibility to enter group statistics instead of raw data.

Author(s)

Mathijs Deen

Examples

library(MASS)
set.seed(1)
ds <- mvrnorm(n=50, mu = c(50,55),
              Sigma = matrix(c(100,0,0,81),
                             ncol = 2),
              empirical = TRUE) |>
  data.frame() |>
  setNames(c("x1","x2"))
t.test(ds$x1, ds$x2)
tTest(x   = ds$x1,
      y   = 55,
      sdy = 9,
      ny  = 50)

Probe willingness to pay

Description

Get the probability of being cost-effective given a certain cost-effectiveness threshold, and vice versa.

Usage

wtp(x, threshold = NULL, probability = NULL)

Arguments

Details

One of the two parameters threshold and probability should be specified.

Value

wtp either the probability or the threshold. If there is no exact match to the given parameter in the bootstrap samples, the result is interpolated.

Author(s)

Mathijs Deen

Examples

CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  wtp(probability = 0.80)
CEA(gnomes, insulationMethod, Costs, diffHATS, 1000, "acorns") |>
  CEAC() |>
  wtp(threshold = 8)