In WillemSleegers/tidystats-v0.3: Create a Tidy Statistics Output File

knitr::opts_chunk$set(echo = TRUE)

# Load packages
library(tidystats)

# Load tidystats results
results <- read_stats("../../tests/testthat/test_results.csv")

# Set the tidystats list in options
options(tidystats_list = results)

t-tests

| Test: | Code: | Output: | |-----------------------------|-------------------------------|---------------------------------| | One sample t-test | report("t_test_one_sample") | r report("t_test_one_sample") | | Two sample t-test | report("t_test_two_sample") | r report("t_test_two_sample") | | Welch's two sample t-test | report("t_test_welch") | r report("t_test_welch") | | Paired t-test | report("t_test_paired") | r report("t_test_paired") |

Single statistic output

| Statistic: | Code: | Output: | |--------------|---------------------------------------------------|-----------------------------------------------------| | t | report("t_test_one_sample", statistic = "t") | r report("t_test_one_sample", s = "t") | | df | report("t_test_one_sample", s = "df") | r report("t_test_one_sample", s = "df") | | p | report("t_test_one_sample", s = "p") | r report("t_test_one_sample", s = "p") | | 9%% CI lower | report("t_test_one_sample", s = "95% CI lower") | r report("t_test_one_sample", s = "95% CI lower") |

Note: the statistic argument can be abbreviated to s.

Correlations

| Test: | Code: | Output: | |----------------------|----------------------------------|------------------------------------| | Pearson correlation | report("correlation_pearson") | r report("correlation_pearson") | | Kendall correlation | report("correlation_kendall") | r report("correlation_kendall") | | Spearman correlation | report("correlation_spearman") | r report("correlation_spearman") |

Single statistic output

| Statistic: | Code: | Output: | |------------|---------------------------------------------|-----------------------------------------------| | r | report("correlation_pearson", s = "r") | r report("correlation_pearson", s = "r") | | r~τ~ | report("correlation_kendall", s = "tau") | r report("correlation_kendall", s = "tau") | | r~s~ | report("correlation_spearman", s = "rho") | r report("correlation_spearman", s = "rho") |

Chi-square tests

| Test: | Code: | Output: | |-----------------------|--------------------------|---------------------------| | Pearson's Chi-squared | report("chi_squared") | r report("chi_squared") |

Single statistic output

| Statistic: | Code: | Output: | |------------|------------------------------------------|--------------------------------------------| | χ² | report("chi_squared", s = "X-squared") | r report("chi_squared", s = "X-squared") |

Wilcoxon rank and signed rank tests

| Test: | Code: | Output: | |---------------------------------------------------|------------------------------------------|--------------------------------------------| | Wilcoxon signed rank test | report("wilcoxon_signed_rank") | r report("wilcoxon_signed_rank") | | Wilcoxon rank sum test | report("wilcoxon_rank_sum") | r report("wilcoxon_rank_sum") | | Wilcoxon rank sum test with CIs | report("wilcoxon_rank_sum_conf") | r report("wilcoxon_rank_sum_conf") | | Wilcoxon rank sum test with continuity correction | report("wilcoxon_rank_sum_continuity") | r report("wilcoxon_rank_sum_continuity") |

Single statistic output

| Statistic: | Code: | Output: | |------------|---------------------------------------------------|-----------------------------------------------------| | V | report("wilcoxon_signed_rank", s = "V") | r report("wilcoxon_signed_rank", s = "V") | | W | report("wilcoxon_rank_sum_continuity", s = "W") | r report("wilcoxon_rank_sum_continuity", s = "W") |

Fisher's exact test for count data

| Test: | Code: | Output: | |--------------------------------------|-------------------------------------|---------------------------------------| | Fisher's exact test for count data | report("fisher_test") | r report("fisher_test") | | Fisher's exact test without CIs | report("fisher_test_no_CI") | r report("fisher_test_no_CI") | | Fisher's test with r > 2 and c > 2 | report("fisher_test_r_by_c") | r report("fisher_test_r_by_c") | | Fisher's test with simulated p-value | report("fisher_test_simulated_p") | r report("fisher_test_simulated_p") | | Fisher's test with hybrid approx. | report("fisher_test_hybrid") | r report("fisher_test_hybrid") |

Single statistic output

| Statistic: | Code: | Output: | |------------|-----------------------------------|-------------------------------------| | OR | report("fisher_test", s = "OR") | r report("fisher_test", s = "OR") |

ANOVA

| Test: | Code: | Output: | |--------------------------------------|-------------------------------------------------------------|---------------------------------------------------------------| | One-way ANOVA | report("aov_one_way", term = "condition") | r report("aov_one_way", term = "condition") | | Two-way ANOVA main effect: condition | report("aov_two_way", term = "condition") | r report("aov_two_way", term = "condition") | | Two-way ANOVA main effect: sex #2 | report("aov_two_way", term = "sex") | r report("aov_two_way", term = "sex") | | Two-way ANOVA interaction #1 | report("aov_two_way_interaction", term = "condition:sex") | r report("aov_two_way_interaction", term = "condition:sex") | | Two-way ANOVA interaction #2 | report("aov_two_way_interaction", term_nr = 3) | r report("aov_two_way_interaction", term_nr = 3) | | ANCOVA | report("aov_ancova", term = "condition") | r report("aov_ancova", term = "condition") | | One-way within-subjects ANOVA | report("aov_one_within", term = "affect") | r report("aov_one_within", term = "affect") | | Mixed ANOVA | report("aov_mixed", term = "affect") | r report("aov_mixed", term = "affect") |

Single statistic output

| Statistic: | Code: | Output: | |-----------------------------------------|--------------------------------------------------------------------------------|----------------------------------------------------------------------------------| | Sum of squares | report("aov_one_way", term = "condition", s = "SS") | r report("aov_one_way", term = "condition", s = "SS") | | F | report("aov_one_way", term = "condition", s = "F") | r report("aov_one_way", term = "condition", s = "F") | | numerator df | report("aov_one_way", term = "condition", s = "df") | r report("aov_one_way", term = "condition", s = "df") | | denominator df | report("aov_one_way", term = "Residuals", s = "df") | r report("aov_one_way", term = "Residuals", s = "df") | | denominator df in within-subjects ANOVA | report("aov_ancova_with_within", group = "ID", term = "Residuals", s = "df") | r report("aov_ancova_with_within", group = "ID", term = "Residuals", s = "df") |

Linear models

| Test: | Code: | Output: | |--------------------------------|--------------------------------------------------------------------------|----------------------------------------------------------------------------| | Regression intercept term | report("lm_simple", term = "(Intercept)") | r report("lm_simple", term = "(Intercept)") | | Regression term #1 | report("lm_simple", term = "conditionmortality salience") | r report("lm_simple", term = "conditionmortality salience") | | Regression term #2 | report("lm_simple", term_nr = 2) | r report("lm_simple", term_nr = 2) | | Regression interaction term #1 | report("lm_interaction", term = "conditionmortality salience:anxiety") | r report("lm_interaction", term = "conditionmortality salience:anxiety") | | Regression interaction term #2 | report("lm_interaction", term_nr = 4) | r report("lm_interaction", term_nr = 4) | | Model fit | report("lm_multiple", group = "model") | r report("lm_multiple", group = "model") | Note: When the term is particularly long and unwieldly, like it is for the interaction term in this example, it may be worthwhile to specify the term_nr argument instead. Of course, this will make the code less readable, so it is recommended to use this conservatively.

Single statistic output

| Statistic: | Code: | Output: | |---------------|---------------------------------------------------|-----------------------------------------------------| | b | report("lm_simple", term_nr = 2, s = "b") | r report("lm_simple", term_nr = 2, s = "b") | | SE | report("lm_simple", term_nr = 2, s = "SE") | r report("lm_simple", term_nr = 2, s = "SE") | | R² | report("lm_multiple", s = "R squared") | r report("lm_multiple", s = "R squared") | | adjusted R² | report("lm_multiple", s = "adjusted R squared") | r report("lm_multiple", s = "adjusted R squared") |

Note: You do not need to specify the group when you want R² because it is a unique statistic. Simply asking for it is enough for tidystats to figure out where it is and report it.

Generalized linear models

| Test: | Code: | Output: | |---------------|--------------------------------------------|----------------------------------------------| | Gaussian term | report("glm_gaussian", term = "Prewt") | r report("glm_gaussian", term = "Prewt") | | Gamma term | report("glm_gamma", term = "log(u)") | r report("glm_gamma", term = "log(u)") | | Poisson term | report("glm_poisson", term = "outcome2") | r report("glm_poisson", term = "outcome2") |

Single statistic output

| Statistic: | Code: | Output: | |---------------------------|-----------------------------------------------------------|-------------------------------------------------------------| | dispersion | report("glm_gaussian", s = "dispersion") | r report("glm_gaussian", s = "dispersion") | | AIC | report("glm_gaussian", s = "AIC") | r report("glm_gaussian", term_nr = 2, s = "SE") | | Fisher Scoring iterations | report("glm_gaussian", s = "Fisher Scoring iterations") | r report("glm_gaussian", s = "Fisher Scoring iterations") |

Linear mixed models

| Test: | Code: | Output: | |--------------------------------------------|--------------------------------------------|----------------------------------------------| | Fixed term: Intercept | report("lme4_lme", term = "(Intercept)") | r report("lme4_lme", term = "(Intercept)") | | Fixed term: Days | report("lme4_lme", term = "Days") | r report("lme4_lme", term = "Days") | | Fixed term: Days (with lmerTest's p-value) | report("lmerTest_lme", term = "Days") | r report("lmerTest_lme", term = "Days") |

Single statistic output

| Statistic: | Code: | Output: | |-----------------------------------------|----------------------------------------------------------|------------------------------------------------------------| | Total N | report("lme4_lme", term = "(Observations)") | r report("lme4_lme", term = "(Observations)", s = "N") | | Total subjects | report("lme4_lme", term = "Subject") | r report("lme4_lme", term = "Subject", s = "N") | | Fixed term: Intercept-Days correlation | report("lme4_lme", term = "(Intercept) - Days" | r report("lme4_lme", term = "(Intercept) - Days") | | Random term: Days variance | report("lme4_lme", term = "Subject - Days", s = "var") | r report("lme4_lme", term = "Subject - Days", s = "var") | | Random term: Intercept-Days correlation | report("lme4_lme", term_nr = 5, s = "r") | r report("lme4_lme", term_nr = 5, s = "r") |

Note: Note that you do not always need to specify the statistic argument to get a single statistic, that is because the term may contain sufficient information for tidystats to figure out what you want to report.

R package: `psych`

| Test: | Code: | Output: | |----------------------------------------------|--------------------------------------------------------------|----------------------------------------------------------------| | corr.test() correlation: rating-complaints | report("psych_correlations", term = "rating - complaints") | r report("psych_correlations", term = "rating - complaints") | | corr.test() correlation: learning-advance | report("psych_correlations", term = "learning - advance") | r report("psych_correlations", term = "learning - advance") |

Single statistic output

| Statistic: | Code: | Output: | |--------------------|--------------------------------------------------|----------------------------------------------------| | Cronbach's alpha | report("psych_alpha", s = "raw_alpha") | r report("psych_alpha", s = "raw_alpha") | | Guttman's Lambda 6 | report("psych_alpha", s = "G6") | r report("psych_alpha", s = "G6") | | ICC1 | report("psych_ICC", group = "ICC1", s = "ICC") | r report("psych_ICC", group = "ICC1", s = "ICC") | | ICC2 | report("psych_ICC", group = "ICC2", s = "ICC") | r report("psych_ICC", group = "ICC2", s = "ICC") | | ICC3 | report("psych_ICC", group = "ICC3", s = "ICC") | r report("psych_ICC", group = "ICC3", s = "ICC") |