NEWS.md
emmeans_test(): restoring grouping variable class (factor) in the final results emmeans_test() (#169)emmeans_test(): “Use of .data in tidyselect expressions was deprecated in tidyselect 1.2.0.”cor_plot() now accepts additional arguments to pass to corrplot() (#66)car::Anova().get_comparisons() now drops unused levels before creating possible comparisons (#67)get_summary_stats() keeps the order of columns specified by the user (#46).two_sample_test() now counts group sizes (n1 and n2) by the number of non-NA values #104
shapiro_test() function. Shapiro_test() throws an error if the input data contains column names “value” or “variable”. This is fixed now (#52).cor_test() function, where there was a tidy evaluation conflict when the input data contains “x” and “y” as column names (#68).dunn_test() documentation is updated to describe the discrepancy between the default behavior of the rstatix::dunn_test() compared to other packages (dunn.test and jamovi). The default of the rstatix::dunn_test() function is to perform a two-sided Dunn test like the well known commercial softwares, such as SPSS and GraphPad. This is not the case for some other R packages (dunn.test and jamovi), where the default is to perform one-sided test (#50).get_summary_stats() handles the user defined probabilities for grouped data (#78)get_n() to extract sample count (n) from statistical test results. - get_description to extract stat test description or name - remove_ns() to remove non-significant rows.add_x_position() to better support different situations (#73).dunn_test() include estimate1 and estimate2 when the argument detailed = TRUE is specified. The estimate1 and estimate2 values represent the mean rank values of the two groups being compared, respectively (#59).cor_spread() doc updated, error is explicitly shown if the input data doesn’t contain the columns “var1”, “var2” and “cor” (#95)emmeans_test() and levene_test() to adapt to broom release 0.7.4 (#89)anova_test() is updated to explain the internal contrast setting (#74).p_mark_significance() works when all p-values are NA. Empty character ("") is returned for NA (#64).rstatix and grouped_anova_test) added to grouped ANOVA test (#61)scales added in the function get_y_position(). If the specified value is “free” or “free_y”, then the step increase of y positions will be calculated by plot panels. Note that, using “free” or “free_y” gives the same result. A global step increase is computed when scales = “fixed” (#56).anova_test() computes now repeated measures ANOVA without error when unused columns are present in the input data frame (#55)stack added in get_y_position() to compute p-values y position for stacked bar plots (#48).wilcox_test(): Now, if detailed = TRUE, an estimate of the location parameter (Only present if argument detailed = TRUE). This corresponds to the pseudomedian (for one-sample case) or to the difference of the location parameter (for two-samples case) (#45).anova_test() function: Changing R default contrast setting (contr.treatment) into orthogonal contrasts (contr.sum) to have comparable results to SPSS when users define the model using formula (@benediktclaus, #40).type = "quantile" of get_summary_stats() works properly (@Boyoron, #39).rstatix and the ggpubr package and makes it easy to program with tidyverse packages using non standard evaluation. - df_select - df_arrange - df_group_by - df_nest_by - df_split_by - df_unite - df_get_var_names - df_label_both - df_label_valuefreq_table() the option na.rm removes only missing values in the variables used to create the frequency table (@JuhlinF, #25).anova_test() (@benediktclaus, #31)games_howell_test() function : the t-statistic is now calculated using the absolute mean difference between groups (@GegznaV, #37).cohens_d() function now supports Hedge’s correction. New argument hedge.correction added . logical indicating whether apply the Hedges correction by multiplying the usual value of Cohen’s d by (N-3)/(N-2.25) (for unpaired t-test) and by (n1-2)/(n1-1.25) for paired t-test; where N is the total size of the two groups being compared (N = n1 + n2) (@IndrajeetPatil, #9).cohens_d() outputs values with directionality. The absolute value is no longer returned. It can now be positive or negative depending on the data (@narunpat, #9).mu is now considered when calculating cohens_d() for one sample t-test (@mllewis, #22).tukey_hsd() now handles situation where minus - symbols are present in factor levels (@IndrajeetPatil, #19).identify_outliers returns a basic data frame instead of tibble when nrow = 0 (for nice printing)detailed added in dunn_test(). If TRUE, then estimate and method columns are shown in the results.prop_test(), pairwise_prop_test() and row_wise_prop_test(). Performs one-sample and two-samples z-test of proportions. Wrappers around the R base function prop.test() but have the advantage of performing pairwise and row-wise z-test of two proportions, the post-hoc tests following a significant chi-square test of homogeneity for 2xc and rx2 contingency tables.fisher_test(), pairwise_fisher_test() and row_wise_fisher_test(): Fisher’s exact test for count data. Wrappers around the R base function fisher.test() but have the advantage of performing pairwise and row-wise fisher tests, the post-hoc tests following a significant chi-square test of homogeneity for 2xc and rx2 contingency tables.chisq_test(), pairwise_chisq_gof_test(), pairwise_chisq_test_against_p() : Chi-square test for count data.binom_test(), pairwise_binom_test(), pairwise_binom_test_against_p() and multinom_test(): performs exact binomial and multinomial tests. Alternative to the chi-square test of goodness-of-fit-test when the sample.counts_to_cases(): converts a contingency table or a data frame of counts into a data frame of individual observations.mcnemar_test() and cochran_qtest() for comparing two ore more related proportions.prop_trend_test(): Performs chi-squared test for trend in proportion. This test is also known as Cochran-Armitage trend test.get_test_label() and get_pwc_label() return expression by defaultget_anova_table() supports now an object of class grouped_anova_test
correction = "none" for repeated measures ANOVANAs are now automatically removed before quantile computation for identifying outliers (@IndrajeetPatil, #10).set_ref_level(), reorder_levels() and make_valid_levels()
model added in the function emmeans_test()
welch_anova_test(): Welch one-Way ANOVA test. A wrapper around the base function stats::oneway.test(). This is is an alternative to the standard one-way ANOVA in the situation where the homogeneity of variance assumption is violated.friedman_effsize(), computes the effect size of Friedman test using the Kendall’s W value.friedman_test(), provides a pipe-friendly framework to perform a Friedman rank sum test, which is the non-parametric alternative to the one-way repeated measures ANOVA test.games_howell_test(): Performs Games-Howell test, which is used to compare all possible combinations of group differences when the assumption of homogeneity of variances is violated.kruskal_effsize() for computing effect size for Kruskal-Wallis test.p_round(), p_format(), p_mark_significant().wilcox_effsize() added for computing effect size (r) for wilcoxon test.get_anova_table() added to extract ANOVA table from anova_test() results. Can apply sphericity correction automatically in the case of within-subject (repeated measures) designs.get_anova_label()
emmeans_test() added for pairwise comparisons of estimated marginal means.comparison removed from tukey_hsd() results (breaking change).n (sample count) added to statistical tests results: t_test(), wilcox_test(), sign_test(), dunn_test() and kruskal_test() (@ShixiangWang, #4).rstatix_test class added to anova_test() resultskruskal_test() is now an object of class rstatix_test that has an attribute named args for holding the test arguments.get_y_position(), y positions and test data are merged now for grouped plots.y.trans added in get_y_position() for y scale transformation.tukey_hsd() results.adjust_pvalue() now supports grouped dataget_pvalue_position added to autocompute p-value positions for plotting significance using ggplot2.get_comparisons() added to create a list of possible pairwise comparisons between groups.dunn_test() added for multiple pairwise comparisons following Kruskal-Wallis test.sign_test() added.get_summary_stats() now supports type = “min”, “max”, “mean” or “median”t_test(), wilcox_test(), dunn_test() and sign_test() are now an object of class rstatix_test that has an attribute named args for holding the test arguments.cohens_d() is now a data frame containing the Cohen’s d and the magnitude.