R/OneStopAnova.R
In AlexandreMillette1989/OneStopAnova: OneStopAnova
Defines functions
OneStopAnova
Documented in
OneStopAnova
#' @title OneStopAnova
#'
#' @description This package generates One-Way and Two-Way Anova with assumptions verification, optional Log and optional Tukey.
#'
#' @usage OneStopAnova(Quantitative = "",
#' Qualitative = "",
#' Qualitative2 = "",
#' var_names = c(Quantitative = "",
#' Qualitative = "",
#' Qualitative2 = ""),
#' Log = TRUE/FALSE,
#' Tukey = TRUE/FALSE)
#'
#' @param Quantitative: The Quantitative variable
#'
#' @param Qualitative: The Qualitative variable
#'
#' @param Qualitative2: The Second Qualitative variable
#'
#' @param var_names: The Names of the Selected Variables
#'
#' @param Log: FALSE by default. If TRUE, Log Transformation of the Quantitative Variable
#'
#' @param Tukey: FALSE by default. If TRUE, Tukey Test on the Selected Variables
#'
#' @import dplyr
#'
#' @import car
#'
#' @import knitr
#'
#' @import nortest
#'
#' @examples
#' OneStopAnova(Quantitative = "calcium$Concentration",
#' Qualitative = "calcium$Trait",
#' Qualitative2 = "calcium$Sexe",
#' var_names = c(Quantitative = "Concentration",
#' Qualitative = "Trait",
#' Qualitative2 = "Sexe"),
#' Log = TRUE,
#' Tukey = TRUE)
#'
#' @export
OneStopAnova = function(Quantitative, Qualitative, Qualitative2, var_names = c(Quantitative = "", Qualitative = "", Qualitative2 = "", ...), Log = c(TRUE, FALSE), Tukey = c(TRUE, FALSE)){
## Var names
var_qualitative = var_names[2]
var_qualitative2 = var_names[3]
var_qualitative_by_qualitative2 = sprintf("%s:%s", var_qualitative, var_qualitative2)
var_names = c(var_qualitative, var_qualitative2, var_qualitative_by_qualitative2)
## Log == FALSE
if((missing(Log))||(Log == FALSE)){
## One-way Anova
if(missing(Qualitative2)){
one_way_anova = aov(Quantitative~Qualitative)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(one_way_anova)~fitted(one_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites", cex.lab = 1.2)
levene = leveneTest(Quantitative ~ Qualitative)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(one_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile", cex.lab = 1.2)
qqline(residuals(one_way_anova))
one_way_anova_residuals = residuals(object = one_way_anova)
anderson_darling = ad.test(one_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "One-Way Anova Anderson-Darling Normality Test \
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "One-Way Anova Anderson-Darling Normality Test \
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform One-Way Anova To DataFrame For Kable Summary
one_way_anova_table = data.frame(unclass(summary(one_way_anova)))
one_way_anova_table = round((one_way_anova_table),3)
one_way_anova_table[, "P_symbols"] = "NA"
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
one_way_anova_table[is.na(one_way_anova_table)] = "NA"
one_way_anova_table = one_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Significance Symbol Loop
for(i in seq_along(one_way_anova_table$P_symbols)){
row = one_way_anova_table$`Pr(>F)`[i]
row_symbol = one_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_table$`Pr(>F)`[i] = row
one_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
one_way_anova_table[one_way_anova_table == "NA"] = ""
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat One-Way ANOVA | No Tukey | No Log
one_way_anova_kable = kable(one_way_anova_table, format = "pandoc", caption = "One-Way Anova Summary")
one_way_anova_summary = list(levene_kable,
anderson_darling_kable,
one_way_anova_kable)
return(one_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
one_way_anova_tukey_qualitative = TukeyHSD(one_way_anova, which = "Qualitative")
one_way_anova_tukey_table = data.frame(unclass(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table = round((one_way_anova_tukey_table ),3)
one_way_anova_tukey_table[, "P_symbols"] = "NA"
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
one_way_anova_tukey_table = one_way_anova_tukey_table %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(one_way_anova_tukey_table) = c(row.names(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table
## Tukey significance Symbol Loop
for(i in seq_along(one_way_anova_tukey_table$Significance)){
row = one_way_anova_tukey_table$Adjusted[i]
row_symbol = one_way_anova_tukey_table$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_tukey_table$Adjusted[i] = row
one_way_anova_tukey_table$Significance[i] = row_symbol
}
colnames(one_way_anova_tukey_table) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "One-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = "Summary"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string4)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string4)
## Listing the summaries
one_way_anova_tukey_print_summary = kable(one_way_anova_tukey_table, format = "pandoc", caption = summary_title_qualitative)
one_way_anova_print_summary = kable(one_way_anova_table, format = "pandoc", caption = summary_title_qualitative2)
one_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
one_way_anova_print_summary,
one_way_anova_tukey_print_summary)
## Résultat One-Way ANOVA | Tukey
return(one_way_anova_print_list)
}
}
## Two-way Anova
else {
two_way_anova = aov(Quantitative ~ Qualitative + Qualitative2 + Qualitative:Qualitative2)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(two_way_anova)~fitted(two_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites", cex.lab = 1.2)
levene = leveneTest(Quantitative ~ Qualitative:Qualitative2)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(two_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile", cex.lab = 1.2)
qqline(residuals(two_way_anova))
two_way_anova_residuals = residuals(object = two_way_anova)
anderson_darling = ad.test(two_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "Two-Way Anova Anderson-Darling Normality Test \
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "Two-Way Anova Anderson-Darling Normality Test \
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform Two-Way Anova To DataFrame For Kable Summary
two_way_anova_table = data.frame(unclass(summary(two_way_anova)))
two_way_anova_table = round((two_way_anova_table),3)
two_way_anova_table[, "P_symbols"] = "NA"
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
two_way_anova_table[is.na(two_way_anova_table)] = "NA"
two_way_anova_table = two_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(two_way_anova_table) = c(var_names[1],var_names[2],var_names[3], "Residuals")
## Significance Symbol Loop
for(i in seq_along(two_way_anova_table$P_symbols)){
row = two_way_anova_table$`Pr(>F)`[i]
row_symbol = two_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_table$`Pr(>F)`[i] = row
two_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
two_way_anova_table[two_way_anova_table == "NA"] = ""
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat Two-Way ANOVA | No Tukey | No Log
two_way_anova_kable = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary")
two_way_anova_summary = list(levene_kable,
anderson_darling_kable,
two_way_anova_kable)
return(two_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
# Qualitative 1
two_way_anova_tukey_qualitative = TukeyHSD(two_way_anova, which = "Qualitative")
two_way_anova_tukey_table_qualitative1 = data.frame(unclass(two_way_anova_tukey_qualitative$Qualitative))
two_way_anova_tukey_table_qualitative1 = round((two_way_anova_tukey_table_qualitative1),3)
two_way_anova_tukey_table_qualitative1[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative1 = two_way_anova_tukey_table_qualitative1 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative1) = c(row.names(two_way_anova_tukey_qualitative$Qualitative))
for(i in seq_along(two_way_anova_tukey_table_qualitative1$Significance)){
row = two_way_anova_tukey_table_qualitative1$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative1$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative1$Adjusted[i] = row
two_way_anova_tukey_table_qualitative1$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative 2
two_way_anova_tukey_qualitative2 = TukeyHSD(two_way_anova, which = "Qualitative2")
two_way_anova_tukey_table_qualitative2 = data.frame(unclass(two_way_anova_tukey_qualitative2$Qualitative2))
two_way_anova_tukey_table_qualitative2 = round((two_way_anova_tukey_table_qualitative2),3)
two_way_anova_tukey_table_qualitative2[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative2 = two_way_anova_tukey_table_qualitative2 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative2) = c(row.names(two_way_anova_tukey_qualitative2$Qualitative2))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative2$Significance)){
row = two_way_anova_tukey_table_qualitative2$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative2$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative2$Adjusted[i] = row
two_way_anova_tukey_table_qualitative2$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative1:Qualitative2
two_way_anova_tukey_qualitative_pair = TukeyHSD(two_way_anova, which = "Qualitative:Qualitative2")
two_way_anova_tukey_table_qualitative_pair = data.frame(unclass(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
two_way_anova_tukey_table_qualitative_pair = round((two_way_anova_tukey_table_qualitative_pair),3)
two_way_anova_tukey_table_qualitative_pair[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative_pair= two_way_anova_tukey_table_qualitative_pair %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative_pair) = c(row.names(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative_pair$Significance)){
row = two_way_anova_tukey_table_qualitative_pair$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative_pair$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative_pair$Adjusted[i] = row
two_way_anova_tukey_table_qualitative_pair$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "Two-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = var_names[3]
summary_title_string5 = "Summary"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string5)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string5)
summary_title_qualitative_pair = paste(summary_title_string1, summary_title_string4, summary_title_string5)
## Listing the summaries
two_way_anova_tukey_print_summary1 = kable(two_way_anova_tukey_table_qualitative1, format = "pandoc", caption = summary_title_qualitative)
two_way_anova_tukey_print_summary2 = kable(two_way_anova_tukey_table_qualitative2, format = "pandoc", caption = summary_title_qualitative2)
two_way_anova_tukey_print_summary_pair = kable(two_way_anova_tukey_table_qualitative_pair, format = "pandoc", caption = summary_title_qualitative_pair)
two_way_anova_print_summary = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary")
two_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
two_way_anova_print_summary,
two_way_anova_tukey_print_summary1,
two_way_anova_tukey_print_summary2,
two_way_anova_tukey_print_summary_pair)
## Résultat Two-Way ANOVA | Tukey | No Log
return(two_way_anova_print_list)
}
}
}
else{
## Log == TRUE
if(Log == TRUE){
Quantitative_Log = log(Quantitative)
## One-Way Anova
if(missing(Qualitative2)){
one_way_anova = aov(Quantitative_Log~Qualitative)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(one_way_anova)~fitted(one_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites (Log)", cex.lab = 1.2)
levene = leveneTest(Quantitative_Log ~ Qualitative)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(one_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile (Log)", cex.lab = 1.2)
qqline(residuals(one_way_anova))
one_way_anova_residuals = residuals(object = one_way_anova)
anderson_darling = ad.test(one_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "One-Way Anova Anderson-Darling Normality Test (Log)\
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "One-Way Anova Anderson-Darling Normality Test (Log)\
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform One-Way Anova To DataFrame For Kable Summary
one_way_anova_table = data.frame(unclass(summary(one_way_anova)))
one_way_anova_table = round((one_way_anova_table),3)
one_way_anova_table[, "P_symbols"] = "NA"
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
one_way_anova_table[is.na(one_way_anova_table)] = "NA"
one_way_anova_table = one_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Significance Symbol Loop
for(i in seq_along(one_way_anova_table$P_symbols)){
row = one_way_anova_table$`Pr(>F)`[i]
row_symbol = one_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_table$`Pr(>F)`[i] = row
one_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
one_way_anova_table[one_way_anova_table == "NA"] = ""
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat One-Way ANOVA | No Tukey | Log
one_way_anova_kable = kable(one_way_anova_table, format = "pandoc", caption = "One-Way Anova Summary (Log)")
one_way_anova_summary = list(levene_kable,
anderson_darling_kable,
one_way_anova_kable)
return(one_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
one_way_anova_tukey_qualitative = TukeyHSD(one_way_anova, which = "Qualitative")
one_way_anova_tukey_table = data.frame(unclass(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table = round((one_way_anova_tukey_table ),3)
one_way_anova_tukey_table[, "P_symbols"] = "NA"
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
one_way_anova_tukey_table = one_way_anova_tukey_table %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(one_way_anova_tukey_table) = c(row.names(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table
## Tukey significance Symbol Loop
for(i in seq_along(one_way_anova_tukey_table$Significance)){
row = one_way_anova_tukey_table$Adjusted[i]
row_symbol = one_way_anova_tukey_table$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_tukey_table$Adjusted[i] = row
one_way_anova_tukey_table$Significance[i] = row_symbol
}
colnames(one_way_anova_tukey_table) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "One-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = "Summary (Log)"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string4)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string4)
## Listing the summaries
one_way_anova_tukey_print_summary = kable(one_way_anova_tukey_table, format = "pandoc", caption = summary_title_qualitative)
one_way_anova_print_summary = kable(one_way_anova_table, format = "pandoc", caption = summary_title_qualitative2)
one_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
one_way_anova_print_summary,
one_way_anova_tukey_print_summary)
## Résultat One-Way ANOVA | Tukey | Log
return(one_way_anova_print_list)
}
}
else {
## Two-Way Anova
two_way_anova = aov(Quantitative_Log ~ Qualitative + Qualitative2 + Qualitative:Qualitative2)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(two_way_anova)~fitted(two_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites (Log)", cex.lab = 1.2)
levene = leveneTest(Quantitative_Log ~ Qualitative:Qualitative2)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(two_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile (Log)", cex.lab = 1.2)
qqline(residuals(two_way_anova))
two_way_anova_residuals = residuals(object = two_way_anova)
anderson_darling = ad.test(two_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "Two-Way Anova Anderson-Darling Normality Test (Log)\
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "Two-Way Anova Anderson-Darling Normality Test (Log)\
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform Two-Way Anova To DataFrame For Kable Summary
two_way_anova_table = data.frame(unclass(summary(two_way_anova)))
two_way_anova_table = round((two_way_anova_table),3)
two_way_anova_table[, "P_symbols"] = "NA"
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
two_way_anova_table[is.na(two_way_anova_table)] = "NA"
two_way_anova_table = two_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(two_way_anova_table) = c(var_names[1],var_names[2],var_names[3], "Residuals")
## Significance Symbol Loop
for(i in seq_along(two_way_anova_table$P_symbols)){
row = two_way_anova_table$`Pr(>F)`[i]
row_symbol = two_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_table$`Pr(>F)`[i] = row
two_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
two_way_anova_table[two_way_anova_table == "NA"] = ""
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat Two-Way ANOVA | No Tukey | Log
two_way_anova_kable = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary (Log)")
two_way_anova_summary = list(levene_kable,
anderson_darling_kable,
two_way_anova_kable)
return(two_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
# Qualitative 1
two_way_anova_tukey_qualitative = TukeyHSD(two_way_anova, which = "Qualitative")
two_way_anova_tukey_table_qualitative1 = data.frame(unclass(two_way_anova_tukey_qualitative$Qualitative))
two_way_anova_tukey_table_qualitative1 = round((two_way_anova_tukey_table_qualitative1),3)
two_way_anova_tukey_table_qualitative1[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative1 = two_way_anova_tukey_table_qualitative1 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative1) = c(row.names(two_way_anova_tukey_qualitative$Qualitative))
for(i in seq_along(two_way_anova_tukey_table_qualitative1$Significance)){
row = two_way_anova_tukey_table_qualitative1$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative1$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative1$Adjusted[i] = row
two_way_anova_tukey_table_qualitative1$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative 2
two_way_anova_tukey_qualitative2 = TukeyHSD(two_way_anova, which = "Qualitative2")
two_way_anova_tukey_table_qualitative2 = data.frame(unclass(two_way_anova_tukey_qualitative2$Qualitative2))
two_way_anova_tukey_table_qualitative2 = round((two_way_anova_tukey_table_qualitative2),3)
two_way_anova_tukey_table_qualitative2[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative2 = two_way_anova_tukey_table_qualitative2 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative2) = c(row.names(two_way_anova_tukey_qualitative2$Qualitative2))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative2$Significance)){
row = two_way_anova_tukey_table_qualitative2$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative2$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative2$Adjusted[i] = row
two_way_anova_tukey_table_qualitative2$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative1:Qualitative2
two_way_anova_tukey_qualitative_pair = TukeyHSD(two_way_anova, which = "Qualitative:Qualitative2")
two_way_anova_tukey_table_qualitative_pair = data.frame(unclass(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
two_way_anova_tukey_table_qualitative_pair = round((two_way_anova_tukey_table_qualitative_pair),3)
two_way_anova_tukey_table_qualitative_pair[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative_pair= two_way_anova_tukey_table_qualitative_pair %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative_pair) = c(row.names(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative_pair$Significance)){
row = two_way_anova_tukey_table_qualitative_pair$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative_pair$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative_pair$Adjusted[i] = row
two_way_anova_tukey_table_qualitative_pair$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "Two-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = var_names[3]
summary_title_string5 = "Summary (Log)"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string5)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string5)
summary_title_qualitative_pair = paste(summary_title_string1, summary_title_string4, summary_title_string5)
## Listing the summaries
two_way_anova_tukey_print_summary1 = kable(two_way_anova_tukey_table_qualitative1, format = "pandoc", caption = summary_title_qualitative)
two_way_anova_tukey_print_summary2 = kable(two_way_anova_tukey_table_qualitative2, format = "pandoc", caption = summary_title_qualitative2)
two_way_anova_tukey_print_summary_pair = kable(two_way_anova_tukey_table_qualitative_pair, format = "pandoc", caption = summary_title_qualitative_pair)
two_way_anova_print_summary = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary (Log)")
two_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
two_way_anova_print_summary,
two_way_anova_tukey_print_summary1,
two_way_anova_tukey_print_summary2,
two_way_anova_tukey_print_summary_pair)
## Résultat Two-Way ANOVA | Tukey | Log
return(two_way_anova_print_list)
}
}
}
}
}
AlexandreMillette1989/OneStopAnova documentation built on Aug. 9, 2020, 3:41 p.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.
Defines functions OneStopAnova
Documented in OneStopAnova
#' @title OneStopAnova
#'
#' @description This package generates One-Way and Two-Way Anova with assumptions verification, optional Log and optional Tukey.
#'
#' @usage OneStopAnova(Quantitative = "",
#' Qualitative = "",
#' Qualitative2 = "",
#' var_names = c(Quantitative = "",
#' Qualitative = "",
#' Qualitative2 = ""),
#' Log = TRUE/FALSE,
#' Tukey = TRUE/FALSE)
#'
#' @param Quantitative: The Quantitative variable
#'
#' @param Qualitative: The Qualitative variable
#'
#' @param Qualitative2: The Second Qualitative variable
#'
#' @param var_names: The Names of the Selected Variables
#'
#' @param Log: FALSE by default. If TRUE, Log Transformation of the Quantitative Variable
#'
#' @param Tukey: FALSE by default. If TRUE, Tukey Test on the Selected Variables
#'
#' @import dplyr
#'
#' @import car
#'
#' @import knitr
#'
#' @import nortest
#'
#' @examples
#' OneStopAnova(Quantitative = "calcium$Concentration",
#' Qualitative = "calcium$Trait",
#' Qualitative2 = "calcium$Sexe",
#' var_names = c(Quantitative = "Concentration",
#' Qualitative = "Trait",
#' Qualitative2 = "Sexe"),
#' Log = TRUE,
#' Tukey = TRUE)
#'
#' @export
OneStopAnova = function(Quantitative, Qualitative, Qualitative2, var_names = c(Quantitative = "", Qualitative = "", Qualitative2 = "", ...), Log = c(TRUE, FALSE), Tukey = c(TRUE, FALSE)){
## Var names
var_qualitative = var_names[2]
var_qualitative2 = var_names[3]
var_qualitative_by_qualitative2 = sprintf("%s:%s", var_qualitative, var_qualitative2)
var_names = c(var_qualitative, var_qualitative2, var_qualitative_by_qualitative2)
## Log == FALSE
if((missing(Log))||(Log == FALSE)){
## One-way Anova
if(missing(Qualitative2)){
one_way_anova = aov(Quantitative~Qualitative)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(one_way_anova)~fitted(one_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites", cex.lab = 1.2)
levene = leveneTest(Quantitative ~ Qualitative)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(one_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile", cex.lab = 1.2)
qqline(residuals(one_way_anova))
one_way_anova_residuals = residuals(object = one_way_anova)
anderson_darling = ad.test(one_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "One-Way Anova Anderson-Darling Normality Test \
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "One-Way Anova Anderson-Darling Normality Test \
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform One-Way Anova To DataFrame For Kable Summary
one_way_anova_table = data.frame(unclass(summary(one_way_anova)))
one_way_anova_table = round((one_way_anova_table),3)
one_way_anova_table[, "P_symbols"] = "NA"
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
one_way_anova_table[is.na(one_way_anova_table)] = "NA"
one_way_anova_table = one_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Significance Symbol Loop
for(i in seq_along(one_way_anova_table$P_symbols)){
row = one_way_anova_table$`Pr(>F)`[i]
row_symbol = one_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_table$`Pr(>F)`[i] = row
one_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
one_way_anova_table[one_way_anova_table == "NA"] = ""
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat One-Way ANOVA | No Tukey | No Log
one_way_anova_kable = kable(one_way_anova_table, format = "pandoc", caption = "One-Way Anova Summary")
one_way_anova_summary = list(levene_kable,
anderson_darling_kable,
one_way_anova_kable)
return(one_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
one_way_anova_tukey_qualitative = TukeyHSD(one_way_anova, which = "Qualitative")
one_way_anova_tukey_table = data.frame(unclass(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table = round((one_way_anova_tukey_table ),3)
one_way_anova_tukey_table[, "P_symbols"] = "NA"
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
one_way_anova_tukey_table = one_way_anova_tukey_table %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(one_way_anova_tukey_table) = c(row.names(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table
## Tukey significance Symbol Loop
for(i in seq_along(one_way_anova_tukey_table$Significance)){
row = one_way_anova_tukey_table$Adjusted[i]
row_symbol = one_way_anova_tukey_table$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_tukey_table$Adjusted[i] = row
one_way_anova_tukey_table$Significance[i] = row_symbol
}
colnames(one_way_anova_tukey_table) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "One-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = "Summary"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string4)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string4)
## Listing the summaries
one_way_anova_tukey_print_summary = kable(one_way_anova_tukey_table, format = "pandoc", caption = summary_title_qualitative)
one_way_anova_print_summary = kable(one_way_anova_table, format = "pandoc", caption = summary_title_qualitative2)
one_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
one_way_anova_print_summary,
one_way_anova_tukey_print_summary)
## Résultat One-Way ANOVA | Tukey
return(one_way_anova_print_list)
}
}
## Two-way Anova
else {
two_way_anova = aov(Quantitative ~ Qualitative + Qualitative2 + Qualitative:Qualitative2)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(two_way_anova)~fitted(two_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites", cex.lab = 1.2)
levene = leveneTest(Quantitative ~ Qualitative:Qualitative2)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) \
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(two_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile", cex.lab = 1.2)
qqline(residuals(two_way_anova))
two_way_anova_residuals = residuals(object = two_way_anova)
anderson_darling = ad.test(two_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "Two-Way Anova Anderson-Darling Normality Test \
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "Two-Way Anova Anderson-Darling Normality Test \
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform Two-Way Anova To DataFrame For Kable Summary
two_way_anova_table = data.frame(unclass(summary(two_way_anova)))
two_way_anova_table = round((two_way_anova_table),3)
two_way_anova_table[, "P_symbols"] = "NA"
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
two_way_anova_table[is.na(two_way_anova_table)] = "NA"
two_way_anova_table = two_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(two_way_anova_table) = c(var_names[1],var_names[2],var_names[3], "Residuals")
## Significance Symbol Loop
for(i in seq_along(two_way_anova_table$P_symbols)){
row = two_way_anova_table$`Pr(>F)`[i]
row_symbol = two_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_table$`Pr(>F)`[i] = row
two_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
two_way_anova_table[two_way_anova_table == "NA"] = ""
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat Two-Way ANOVA | No Tukey | No Log
two_way_anova_kable = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary")
two_way_anova_summary = list(levene_kable,
anderson_darling_kable,
two_way_anova_kable)
return(two_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
# Qualitative 1
two_way_anova_tukey_qualitative = TukeyHSD(two_way_anova, which = "Qualitative")
two_way_anova_tukey_table_qualitative1 = data.frame(unclass(two_way_anova_tukey_qualitative$Qualitative))
two_way_anova_tukey_table_qualitative1 = round((two_way_anova_tukey_table_qualitative1),3)
two_way_anova_tukey_table_qualitative1[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative1 = two_way_anova_tukey_table_qualitative1 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative1) = c(row.names(two_way_anova_tukey_qualitative$Qualitative))
for(i in seq_along(two_way_anova_tukey_table_qualitative1$Significance)){
row = two_way_anova_tukey_table_qualitative1$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative1$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative1$Adjusted[i] = row
two_way_anova_tukey_table_qualitative1$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative 2
two_way_anova_tukey_qualitative2 = TukeyHSD(two_way_anova, which = "Qualitative2")
two_way_anova_tukey_table_qualitative2 = data.frame(unclass(two_way_anova_tukey_qualitative2$Qualitative2))
two_way_anova_tukey_table_qualitative2 = round((two_way_anova_tukey_table_qualitative2),3)
two_way_anova_tukey_table_qualitative2[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative2 = two_way_anova_tukey_table_qualitative2 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative2) = c(row.names(two_way_anova_tukey_qualitative2$Qualitative2))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative2$Significance)){
row = two_way_anova_tukey_table_qualitative2$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative2$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative2$Adjusted[i] = row
two_way_anova_tukey_table_qualitative2$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative1:Qualitative2
two_way_anova_tukey_qualitative_pair = TukeyHSD(two_way_anova, which = "Qualitative:Qualitative2")
two_way_anova_tukey_table_qualitative_pair = data.frame(unclass(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
two_way_anova_tukey_table_qualitative_pair = round((two_way_anova_tukey_table_qualitative_pair),3)
two_way_anova_tukey_table_qualitative_pair[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative_pair= two_way_anova_tukey_table_qualitative_pair %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative_pair) = c(row.names(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative_pair$Significance)){
row = two_way_anova_tukey_table_qualitative_pair$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative_pair$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative_pair$Adjusted[i] = row
two_way_anova_tukey_table_qualitative_pair$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "Two-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = var_names[3]
summary_title_string5 = "Summary"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string5)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string5)
summary_title_qualitative_pair = paste(summary_title_string1, summary_title_string4, summary_title_string5)
## Listing the summaries
two_way_anova_tukey_print_summary1 = kable(two_way_anova_tukey_table_qualitative1, format = "pandoc", caption = summary_title_qualitative)
two_way_anova_tukey_print_summary2 = kable(two_way_anova_tukey_table_qualitative2, format = "pandoc", caption = summary_title_qualitative2)
two_way_anova_tukey_print_summary_pair = kable(two_way_anova_tukey_table_qualitative_pair, format = "pandoc", caption = summary_title_qualitative_pair)
two_way_anova_print_summary = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary")
two_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
two_way_anova_print_summary,
two_way_anova_tukey_print_summary1,
two_way_anova_tukey_print_summary2,
two_way_anova_tukey_print_summary_pair)
## Résultat Two-Way ANOVA | Tukey | No Log
return(two_way_anova_print_list)
}
}
}
else{
## Log == TRUE
if(Log == TRUE){
Quantitative_Log = log(Quantitative)
## One-Way Anova
if(missing(Qualitative2)){
one_way_anova = aov(Quantitative_Log~Qualitative)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(one_way_anova)~fitted(one_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites (Log)", cex.lab = 1.2)
levene = leveneTest(Quantitative_Log ~ Qualitative)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "One-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(one_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile (Log)", cex.lab = 1.2)
qqline(residuals(one_way_anova))
one_way_anova_residuals = residuals(object = one_way_anova)
anderson_darling = ad.test(one_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "One-Way Anova Anderson-Darling Normality Test (Log)\
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "One-Way Anova Anderson-Darling Normality Test (Log)\
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform One-Way Anova To DataFrame For Kable Summary
one_way_anova_table = data.frame(unclass(summary(one_way_anova)))
one_way_anova_table = round((one_way_anova_table),3)
one_way_anova_table[, "P_symbols"] = "NA"
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
one_way_anova_table[is.na(one_way_anova_table)] = "NA"
one_way_anova_table = one_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Significance Symbol Loop
for(i in seq_along(one_way_anova_table$P_symbols)){
row = one_way_anova_table$`Pr(>F)`[i]
row_symbol = one_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_table$`Pr(>F)`[i] = row
one_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
one_way_anova_table[one_way_anova_table == "NA"] = ""
colnames(one_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
rownames(one_way_anova_table) = c(var_names[1], "Residuals")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat One-Way ANOVA | No Tukey | Log
one_way_anova_kable = kable(one_way_anova_table, format = "pandoc", caption = "One-Way Anova Summary (Log)")
one_way_anova_summary = list(levene_kable,
anderson_darling_kable,
one_way_anova_kable)
return(one_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
one_way_anova_tukey_qualitative = TukeyHSD(one_way_anova, which = "Qualitative")
one_way_anova_tukey_table = data.frame(unclass(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table = round((one_way_anova_tukey_table ),3)
one_way_anova_tukey_table[, "P_symbols"] = "NA"
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
one_way_anova_tukey_table = one_way_anova_tukey_table %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(one_way_anova_tukey_table) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(one_way_anova_tukey_table) = c(row.names(one_way_anova_tukey_qualitative$Qualitative))
one_way_anova_tukey_table
## Tukey significance Symbol Loop
for(i in seq_along(one_way_anova_tukey_table$Significance)){
row = one_way_anova_tukey_table$Adjusted[i]
row_symbol = one_way_anova_tukey_table$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
one_way_anova_tukey_table$Adjusted[i] = row
one_way_anova_tukey_table$Significance[i] = row_symbol
}
colnames(one_way_anova_tukey_table) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "One-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = "Summary (Log)"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string4)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string4)
## Listing the summaries
one_way_anova_tukey_print_summary = kable(one_way_anova_tukey_table, format = "pandoc", caption = summary_title_qualitative)
one_way_anova_print_summary = kable(one_way_anova_table, format = "pandoc", caption = summary_title_qualitative2)
one_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
one_way_anova_print_summary,
one_way_anova_tukey_print_summary)
## Résultat One-Way ANOVA | Tukey | Log
return(one_way_anova_print_list)
}
}
else {
## Two-Way Anova
two_way_anova = aov(Quantitative_Log ~ Qualitative + Qualitative2 + Qualitative:Qualitative2)
par(mfrow = c(1,2))
## 1. Supposition homoscédasticité
plot(residuals(two_way_anova)~fitted(two_way_anova), ylab = "Résidus", xlab = "Valeurs prédites", main = "Résidus vs valeurs prédites (Log)", cex.lab = 1.2)
levene = leveneTest(Quantitative_Log ~ Qualitative:Qualitative2)
levene = round((levene),3)
options(knitr.kable.NA = "")
levene = levene %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == 0, 0.001, `Pr(>F)`))
if(levene$`Pr(>F)`[1] <= 0.05){
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
WARNING: Evidence suggests that the variance across groups is statistically significantly different. \
WARNING: We cannot assume the homogeneity of variances in the different treatment groups."
} else {
levene_title = "Two-Way Anova Levene's Test for Homogeneity of Variance (center = median) (Log)\
\
There is no evidence to suggest that the variance across groups is statistically significantly different. \
We can assume the homogeneity of variances in the different treatment groups."
}
levene_kable = kable(levene, caption = levene_title,
format = "pandoc")
## 2. Supposition de normalité des résidus
qqnorm(residuals(two_way_anova), ylab = "Quantiles observés", xlab = "Quantiles théoriques", main = "Graphique quantile-quantile (Log)", cex.lab = 1.2)
qqline(residuals(two_way_anova))
two_way_anova_residuals = residuals(object = two_way_anova)
anderson_darling = ad.test(two_way_anova_residuals)
anderson_darling_table = cbind(anderson_darling$statistic, anderson_darling$p.value)
colnames(anderson_darling_table) = c("Test Statistic", "p")
rownames(anderson_darling_table) = c()
anderson_darling_table = round((anderson_darling_table),3)
anderson_darling_table = as.data.frame(anderson_darling_table)
anderson_darling_table = anderson_darling_table %>%
mutate(
p = if_else(p == 0, 0.001, p))
colnames(anderson_darling_table) = c("Test Statistic", "P-Value")
if(anderson_darling$p.value <= 0.05){
ad_title = "Two-Way Anova Anderson-Darling Normality Test (Log)\
\
WARNING: We have sufficient evidence to reject the null hypothesis.
WARNING: It is safe to say that the data doesn't follow a normal distribution."
} else {
ad_title = "Two-Way Anova Anderson-Darling Normality Test (Log)\
\
We do not have sufficient evidence to reject the null hypothesis.
It is safe to say that the data follows a normal distribution."
}
anderson_darling_kable = kable(anderson_darling_table, caption = ad_title, format = "pandoc")
## Transform Two-Way Anova To DataFrame For Kable Summary
two_way_anova_table = data.frame(unclass(summary(two_way_anova)))
two_way_anova_table = round((two_way_anova_table),3)
two_way_anova_table[, "P_symbols"] = "NA"
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "P_symbols")
two_way_anova_table[is.na(two_way_anova_table)] = "NA"
two_way_anova_table = two_way_anova_table %>%
mutate(
`Pr(>F)` = if_else(`Pr(>F)` == "0", "0.001", `Pr(>F)`))
rownames(two_way_anova_table) = c(var_names[1],var_names[2],var_names[3], "Residuals")
## Significance Symbol Loop
for(i in seq_along(two_way_anova_table$P_symbols)){
row = two_way_anova_table$`Pr(>F)`[i]
row_symbol = two_way_anova_table$P_symbols[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_table$`Pr(>F)`[i] = row
two_way_anova_table$P_symbols[i] = row_symbol
}
## Rearrange Table
two_way_anova_table[two_way_anova_table == "NA"] = ""
colnames(two_way_anova_table) = c("Df", "Sum Sq", "Mean Sq", "F Value", "Pr(>F)", "Significance Symbol")
## Tukey == FALSE
if(missing(Tukey)||(Tukey == FALSE)){
## Résultat Two-Way ANOVA | No Tukey | Log
two_way_anova_kable = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary (Log)")
two_way_anova_summary = list(levene_kable,
anderson_darling_kable,
two_way_anova_kable)
return(two_way_anova_summary)
}
## Tukey == TRUE
if(Tukey == TRUE){
# Qualitative 1
two_way_anova_tukey_qualitative = TukeyHSD(two_way_anova, which = "Qualitative")
two_way_anova_tukey_table_qualitative1 = data.frame(unclass(two_way_anova_tukey_qualitative$Qualitative))
two_way_anova_tukey_table_qualitative1 = round((two_way_anova_tukey_table_qualitative1),3)
two_way_anova_tukey_table_qualitative1[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative1 = two_way_anova_tukey_table_qualitative1 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative1) = c(row.names(two_way_anova_tukey_qualitative$Qualitative))
for(i in seq_along(two_way_anova_tukey_table_qualitative1$Significance)){
row = two_way_anova_tukey_table_qualitative1$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative1$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative1$Adjusted[i] = row
two_way_anova_tukey_table_qualitative1$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative1) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative 2
two_way_anova_tukey_qualitative2 = TukeyHSD(two_way_anova, which = "Qualitative2")
two_way_anova_tukey_table_qualitative2 = data.frame(unclass(two_way_anova_tukey_qualitative2$Qualitative2))
two_way_anova_tukey_table_qualitative2 = round((two_way_anova_tukey_table_qualitative2),3)
two_way_anova_tukey_table_qualitative2[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative2 = two_way_anova_tukey_table_qualitative2 %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative2) = c(row.names(two_way_anova_tukey_qualitative2$Qualitative2))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative2$Significance)){
row = two_way_anova_tukey_table_qualitative2$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative2$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative2$Adjusted[i] = row
two_way_anova_tukey_table_qualitative2$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative2) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
# Qualitative1:Qualitative2
two_way_anova_tukey_qualitative_pair = TukeyHSD(two_way_anova, which = "Qualitative:Qualitative2")
two_way_anova_tukey_table_qualitative_pair = data.frame(unclass(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
two_way_anova_tukey_table_qualitative_pair = round((two_way_anova_tukey_table_qualitative_pair),3)
two_way_anova_tukey_table_qualitative_pair[, "P_symbols"] = "NA"
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
two_way_anova_tukey_table_qualitative_pair= two_way_anova_tukey_table_qualitative_pair %>%
mutate(
Adjusted = if_else(Adjusted == 0, 0.001, Adjusted))
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower", "Upper", "Adjusted", "Significance")
rownames(two_way_anova_tukey_table_qualitative_pair) = c(row.names(two_way_anova_tukey_qualitative_pair$`Qualitative:Qualitative2`))
## Tukey significance Symbol Loop
for(i in seq_along(two_way_anova_tukey_table_qualitative_pair$Significance)){
row = two_way_anova_tukey_table_qualitative_pair$Adjusted[i]
row_symbol = two_way_anova_tukey_table_qualitative_pair$Significance[i]
if(row <= 0.001){
row_symbol = paste0("***")
}else if(row <= 0.01){
row_symbol = paste0("**")
}else if(row <= 0.05){
row_symbol = paste0("*")
} else if(row <= 0.10){
row_symbol = paste0(".")
} else if(row <= ""){
row_symbol = paste0(" ")
} else
row_symbol = paste0(" ")
two_way_anova_tukey_table_qualitative_pair$Adjusted[i] = row
two_way_anova_tukey_table_qualitative_pair$Significance[i] = row_symbol
}
colnames(two_way_anova_tukey_table_qualitative_pair) = c("Difference", "Lower 95", "Upper 95", "Adjusted P-Value", "Significance Symbol")
## Summary String Title
summary_title_string1 = "Two-Way Anova Tukey"
summary_title_string2 = var_names[1]
summary_title_string3 = var_names[2]
summary_title_string4 = var_names[3]
summary_title_string5 = "Summary (Log)"
summary_title_qualitative = paste(summary_title_string1, summary_title_string2, summary_title_string5)
summary_title_qualitative2 = paste(summary_title_string1, summary_title_string3, summary_title_string5)
summary_title_qualitative_pair = paste(summary_title_string1, summary_title_string4, summary_title_string5)
## Listing the summaries
two_way_anova_tukey_print_summary1 = kable(two_way_anova_tukey_table_qualitative1, format = "pandoc", caption = summary_title_qualitative)
two_way_anova_tukey_print_summary2 = kable(two_way_anova_tukey_table_qualitative2, format = "pandoc", caption = summary_title_qualitative2)
two_way_anova_tukey_print_summary_pair = kable(two_way_anova_tukey_table_qualitative_pair, format = "pandoc", caption = summary_title_qualitative_pair)
two_way_anova_print_summary = kable(two_way_anova_table, format = "pandoc", caption = "Two-Way Anova Summary (Log)")
two_way_anova_print_list = list(levene_kable,
anderson_darling_kable,
two_way_anova_print_summary,
two_way_anova_tukey_print_summary1,
two_way_anova_tukey_print_summary2,
two_way_anova_tukey_print_summary_pair)
## Résultat Two-Way ANOVA | Tukey | Log
return(two_way_anova_print_list)
}
}
}
}
}
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.