fixes

DESCRIPTION

@@ -27,17 +27,14 @@ Imports:
     performance
 Suggests:
     brms,
-    broom,
     car,
     coin,
     ggplot2,
-    graphics,
     MASS,
     pscl,
     pwr,
     sjPlot,
     survey,
-    rstan,
     testthat
 URL: https://strengejacke.github.io/sjstats/
 BugReports: https://github.com/strengejacke/sjstats/issues

NAMESPACE

@@ -114,7 +114,6 @@ importFrom(datawizard,weighted_sd)
 importFrom(insight,export_table)
 importFrom(insight,format_p)
 importFrom(insight,format_value)
-importFrom(insight,get_response)
 importFrom(insight,link_inverse)
 importFrom(insight,print_color)
 importFrom(performance,mse)

R/S3-methods.R

@@ -63,8 +63,8 @@ tidy_svyglm.nb <- function(x, digits = 4, v_se = c("robust", "model")) {
     estimate = round(est, digits),
     irr = round(exp(est), digits),
     std.error = round(se, digits),
-    conf.low = round(exp(est - stats::qnorm(.975) * se), digits),
-    conf.high = round(exp(est + stats::qnorm(.975) * se), digits),
+    conf.low = round(exp(est - stats::qnorm(0.975) * se), digits),
+    conf.high = round(exp(est + stats::qnorm(0.975) * se), digits),
     p.value = round(2 * stats::pnorm(abs(est / se), lower.tail = FALSE), digits)
   )
 }
@@ -265,9 +265,13 @@ print.sj_resample <- function(x, ...) {
   else
     id10 <- x$id
 
-  cat("<", paste0("id's of resample [", prettyNum(nrow(x$data), big.mark = ","), " x ",
-                  prettyNum(ncol(x$data), big.mark = ","), "]"), "> ",
-      paste(id10, collapse = ", "), "\n", sep = "")
+  cat("<", paste0(
+    "id's of resample [", prettyNum(nrow(x$data), big.mark = ","), " x ",
+    prettyNum(ncol(x$data), big.mark = ","), "]"
+  ), "> ",
+  toString(id10), "\n",
+  sep = ""
+  )
 }
 
 

R/anova_stats.R

@@ -15,7 +15,7 @@
 #'   \cr \cr
 #'   Tippey K, Longnecker MT (2016): An Ad Hoc Method for Computing Pseudo-Effect Size for Mixed Model.
 #'
-#' @examples
+#' @examplesIf requireNamespace("car")
 #' # load sample data
 #' data(efc)
 #'
@@ -24,9 +24,7 @@
 #'   c12hour ~ as.factor(e42dep) + as.factor(c172code) + c160age,
 #'   data = efc
 #' )
-#' \dontrun{
 #' anova_stats(car::Anova(fit, type = 2))
-#' }
 #' @export
 anova_stats <- function(model, digits = 3) {
   insight::check_if_installed("pwr")
@@ -97,7 +95,7 @@ aov_stat <- function(model, type) {
 
 
 aov_stat_summary <- function(model) {
-  insight::check_if_installed("broom")
+  insight::check_if_installed("parameters")
   # check if we have a mixed model
   mm <- is_merMod(model)
   ori.model <- model
@@ -108,7 +106,7 @@ aov_stat_summary <- function(model) {
     model <- stats::anova(model)
 
   # get summary table
-  aov.sum <- as.data.frame(broom::tidy(model))
+  aov.sum <- insight::standardize_names(as.data.frame(parameters::model_parameters(model)), style = "broom")
 
   # for mixed models, add information on residuals
   if (mm) {

R/auto_prior.R

@@ -34,20 +34,16 @@
 #'   formula used in \code{brms::brm()} must be rewritten to something like
 #'   \code{y ~ 0 + intercept ...}, see \code{\link[brms]{set_prior}}.
 #'
-#' @examples
-#' library(sjmisc)
+#' @examplesIf requireNamespace("brms")
 #' data(efc)
 #' efc$c172code <- as.factor(efc$c172code)
-#' efc$c161sex <- to_label(efc$c161sex)
+#' efc$c161sex <- as.factor(efc$c161sex)
 #'
 #' mf <- formula(neg_c_7 ~ c161sex + c160age + c172code)
-#'
-#' if (requireNamespace("brms", quietly = TRUE))
-#'   auto_prior(mf, efc, TRUE)
+#' auto_prior(mf, efc, TRUE)
 #'
 #' ## compare to
-#' # library(rstanarm)
-#' # m <- stan_glm(mf, data = efc, chains = 2, iter = 200)
+#' # m <- rstanarm::stan_glm(mf, data = efc, chains = 2, iter = 200)
 #' # ps <- prior_summary(m)
 #' # ps$prior_intercept$adjusted_scale
 #' # ps$prior$adjusted_scale
@@ -59,17 +55,13 @@
 #' # add informative priors
 #' mf <- formula(neg_c_7 ~ c161sex + c172code)
 #'
-#' if (requireNamespace("brms", quietly = TRUE)) {
-#'   auto_prior(mf, efc, TRUE) +
-#'     brms::prior(normal(.1554, 40), class = "b", coef = "c160age")
-#' }
+#' auto_prior(mf, efc, TRUE) +
+#'   brms::prior(normal(.1554, 40), class = "b", coef = "c160age")
 #'
 #' # example with binary response
 #' efc$neg_c_7d <- ifelse(efc$neg_c_7 < median(efc$neg_c_7, na.rm = TRUE), 0, 1)
 #' mf <- formula(neg_c_7d ~ c161sex + c160age + c172code + e17age)
-#'
-#' if (requireNamespace("brms", quietly = TRUE))
-#'   auto_prior(mf, efc, FALSE)
+#' auto_prior(mf, efc, FALSE)
 #' @export
 auto_prior <- function(formula, data, gaussian, locations = NULL) {
   insight::check_if_installed("brms")

R/boot_ci.R

@@ -42,7 +42,7 @@
 #'
 #' @seealso []`bootstrap()`] to generate nonparametric bootstrap samples.
 #'
-#' @examplesIf getRversion() >= "4.2.0"
+#' @examples
 #' data(efc)
 #' bs <- bootstrap(efc, 100)
 #'

R/cramer.R

@@ -23,19 +23,19 @@ cramers_v.ftable <- function(tab, ...) {
 #' @rdname crosstable_statistics
 #' @export
 cramers_v.formula <- function(formula, data, ci.lvl = NULL, n = 1000, method = c("dist", "quantile"), ...) {
-  terms <- all.vars(formula)
-  tab <- table(data[[terms[1]]], data[[terms[2]]])
+  fterms <- all.vars(formula)
+  tab <- table(data[[fterms[1]]], data[[fterms[2]]])
   method <- match.arg(method)
 
   if (is.null(ci.lvl) || is.na(ci.lvl)) {
     .cramers_v(tab)
   } else {
-    straps <- sjstats::bootstrap(data[terms], n)
+    straps <- sjstats::bootstrap(data[fterms], n)
     tables <- lapply(straps$strap, function(x) {
       dat <- as.data.frame(x)
       table(dat[[1]], dat[[2]])
     })
-    cramers <- sapply(tables, function(x) .cramers_v(x))
+    cramers <- sapply(tables, .cramers_v)
     ci <- boot_ci(cramers, ci.lvl = ci.lvl, method = method)
 
     data_frame(

R/cv.R

@@ -36,9 +36,8 @@ cv <- function(x, ...) {
 }
 
 
-#' @importFrom performance rmse
-#' @importFrom insight get_response
 cv_helper <- function(x) {
+  insight::check_if_installed("performance")
   # check if we have a fitted linear model
   if (inherits(x, c("lm", "lmerMod", "lme", "merModLmerTest")) && !inherits(x, "glm")) {
     # get response

R/find_beta.R

@@ -137,24 +137,24 @@ find_beta2 <- function(x, se, ci, n) {
 
     # need to compute proportion
     x <- x / n
-    p2 <- .95
+    p2 <- 0.95
     x2 <- x + bvar
   }
 
   # for standard errors, we assume a 68% quantile
   if (!missing(se)) {
-    p2 <- .68
+    p2 <- 0.68
     x2 <- x + se
   }
 
   # for CI, we assume a 68% quantile
   if (!missing(ci)) {
-    p2 <- .95
+    p2 <- 0.95
     x2 <- ci
   }
 
   # the probability is assumed to be the median
-  p1 <- .5
+  p1 <- 0.5
   x1 <- x
 
   find_beta(x1, p1, x2, p2)
@@ -182,4 +182,3 @@ find_normal <- function(x1, p1, x2, p2) {
 
   list(mean = mw, sd = stddev)
 }
-