Improved Function Compatibility and Dependency Updates

- Enhanced compatibility of the `report` function for cases when `alpha.name = NULL` and `dist.name = NULL`. - Reduced package dependencies. Removed dependency on packages like `broom`. - Updated package dependencies as follows: **Depends:** - R (>= 3.5.0) - rlang - tibble **Imports:** - ggplot2 - matrixStats - lmerTest - foreach - modeest - vegan - dplyr - pheatmap - tidyr - ggh4x - ape - GUniFrac - scales - stringr - rmarkdown - knitr - pander - tinytex **Suggests:** - ggrepel - parallel - ggprism - aplot - philentropy - forcats - yaml - biomformat - Biostrings
cafferychen777 · Oct 13, 2023 · 23c0b9f · 23c0b9f
1 parent dea61d6
commit 23c0b9f
Show file tree

Hide file tree

Showing 54 changed files with 368 additions and 109 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -23,8 +23,6 @@ Depends:
 Imports: 
     ggplot2,
     matrixStats,
-    parallel,
-    ggrepel,
     lmerTest,
     foreach,
     modeest,
@@ -37,16 +35,17 @@ Imports:
     GUniFrac,
     scales,
     stringr,
-    broom
-Suggests:
     rmarkdown,
     knitr,
+    pander,
+    tinytex
+Suggests:
+    ggrepel,
+    parallel,
     ggprism,
     aplot,
     philentropy,
     forcats,
-    pander,
-    tinytex,
     yaml,
     biomformat,
     Biostrings

diff --git a/NAMESPACE b/NAMESPACE
@@ -97,7 +97,6 @@ export(mStat_update_sample_name)
 export(mStat_validate_data)
 import(foreach)
 import(ggplot2)
-import(ggrepel)
 import(parallel)
 import(rlang)
 import(tibble)

diff --git a/R/generate_alpha_boxplot_long.R b/R/generate_alpha_boxplot_long.R
@@ -102,6 +102,11 @@ generate_alpha_boxplot_long <- function (data.obj,
                                          pdf.wid = 11,
                                          pdf.hei = 8.5,
                                          ...) {
+
+  if (is.null(alpha.name)){
+    return()
+  }
+
   if (!is.null(alpha.obj) &&
       !is(alpha.obj, "list"))
     stop("`alpha.obj` should be a list or NULL.")

diff --git a/R/generate_alpha_boxplot_single.R b/R/generate_alpha_boxplot_single.R
@@ -160,6 +160,11 @@ generate_alpha_boxplot_single <- function (data.obj,
                                            pdf.wid = 11,
                                            pdf.hei = 8.5,
                                            ...) {
+
+  if (is.null(alpha.name)){
+    return()
+  }
+
   if (is.null(alpha.obj)) {
     if (!is_rarefied(data.obj)) {
       message(

diff --git a/R/generate_alpha_change_boxplot_pair.R b/R/generate_alpha_change_boxplot_pair.R
@@ -112,6 +112,11 @@ generate_alpha_change_boxplot_pair <-
            pdf.wid = 11,
            pdf.hei = 8.5,
            ...) {
+
+    if (is.null(alpha.name)){
+      return()
+    }
+
     if (is.null(alpha.obj)) {
       if (!is_rarefied(data.obj)) {
         message(

diff --git a/R/generate_alpha_change_test_pair.R b/R/generate_alpha_change_test_pair.R
@@ -84,10 +84,10 @@
 #' time.var = "time",
 #' alpha.name = c("shannon"),
 #' subject.var = "subject",
-#' group.var = "group",
-#' adj.vars = "sex",
-#' change.base = "1",
-#' alpha.change.func = "absolute change"
+#' group.var = "sex",
+#' adj.vars = "group",
+#' change.base = "2",
+#' alpha.change.func = "log fold change"
 #' )
 #' }
 #' @export
@@ -103,6 +103,10 @@ generate_alpha_change_test_pair <-
            change.base,
            alpha.change.func = "log fold change") {
 
+    if (is.null(alpha.name)){
+      return()
+    }
+
     if (is.null(alpha.obj)) {
       if (!is_rarefied(data.obj)) {
         message(
@@ -201,40 +205,37 @@ generate_alpha_change_test_pair <-
 
       # Run lm and create a coefficient table
       lm.model <- lm(formula, data = combined_alpha)
-      coef.tab <- broom::tidy(summary(lm.model))
-
-      # Rearrange the table
-      coef.tab <-
-        coef.tab %>% dplyr::select(
-          Term = term,
-          Estimate = estimate,
-          Std.Error = std.error,
-          Statistic = statistic,
-          P.Value = p.value
-        )
+      summary <- summary(lm.model)
+      coef.tab <- summary$coefficients %>%
+        as.data.frame() %>%
+        tibble::rownames_to_column(var = "term") %>%
+        rename(
+          term = "Term",
+          `Std. Error` = "Std.Error",
+          `t value` = "Statistic",
+          `Pr(>|t|)` = "P.Value"
+        ) %>% as_tibble()
 
       # Run ANOVA on the model if group.var is multi-categorical
       if (length(unique(combined_alpha[[group.var]])) > 2) {
-        anova.tab <- broom::tidy(anova(lm.model))
-
-        # Rearrange the table and add missing columns
-        anova.tab <- anova.tab %>%
-          dplyr::select(
-            Term = term,
-            Statistic = statistic,
-            df = df,
-            P.Value = p.value
-          ) %>%
-          dplyr::mutate(Estimate = NA, Std.Error = NA)
+        anova <- anova(lm.model)
+        anova.tab <- as.data.frame(anova) %>%
+          rownames_to_column("Term") %>%
+          rename(`Sum Sq` = "sumsq",
+                 `Df` = "df",
+                 `F value` = "Statistic",
+                 `Pr(>F)` = "P.Value") %>%
+          dplyr::mutate(Estimate = NA, Std.Error = NA) %>%
+          as_tibble()
 
         # Reorder the columns to match coef.tab
         anova.tab <- anova.tab %>%
           dplyr::select(
-            Term = term,
-            Estimate = Estimate,
-            Std.Error = Std.Error,
-            Statistic = Statistic,
-            P.Value = P.Value
+            Term,
+            Estimate,
+            Std.Error,
+            Statistic,
+            P.Value
           )
 
         coef.tab <-

diff --git a/R/generate_alpha_spaghettiplot_long.R b/R/generate_alpha_spaghettiplot_long.R
@@ -91,6 +91,10 @@ generate_alpha_spaghettiplot_long <-
            pdf.wid = 11,
            pdf.hei = 8.5,
            ...) {
+    if (is.null(alpha.name)){
+      return()
+    }
+
     # Data validation
     if (!is(data.obj, "list"))
       stop("`data.obj` should be a list.")

diff --git a/R/generate_alpha_test_pair.R b/R/generate_alpha_test_pair.R
@@ -87,7 +87,7 @@ extract_coef <- function(model) {
 #' alpha.name = c("shannon"),
 #' subject.var = "subject",
 #' group.var = "group",
-#' adj.vars = "sex"
+#' adj.vars = NULL
 #' )
 #'
 #' @export
@@ -100,6 +100,11 @@ generate_alpha_test_pair <-
            subject.var,
            group.var,
            adj.vars) {
+
+    if (is.null(alpha.name)){
+      return()
+    }
+
     if (is.null(alpha.obj)) {
       if (!is_rarefied(data.obj)) {
         message(
@@ -187,7 +192,12 @@ generate_alpha_test_pair <-
     if(!is.null(group.var)){
       # Run ANOVA on the model if group.var is multi-categorical
       if (group.levels > 2) {
-        anova.tab <- broom::tidy(anova(lme.model))
+        anova.tab <- anova(lme.model) %>%
+          as.data.frame() %>%
+          rownames_to_column("Term") %>%
+          rename(`F value` = "statistic",
+                 `Pr(>F)` = "p.value") %>%
+          as_tibble()
 
         # Rearrange the table and add missing columns
         anova.tab <- anova.tab %>%
@@ -196,13 +206,13 @@ generate_alpha_test_pair <-
         # Reorder the columns to match coef.tab
         anova.tab <- anova.tab %>%
           dplyr::select(
-            Term = term,
-            Estimate = Estimate,
-            Std.Error = Std.Error,
+            Term,
+            Estimate,
+            Std.Error,
             Statistic = statistic,
             P.Value = p.value
           ) %>%
-          dplyr::filter(Term %in% c(group.var, paste0(time.var, ":", group.var)))
+          dplyr::filter(Term %in% c(group.var, paste0(group.var, ":", time.var)))
 
         coef.tab <-
           rbind(coef.tab, anova.tab) # Append the anova.tab to the coef.tab

diff --git a/R/generate_alpha_test_single.R b/R/generate_alpha_test_single.R
@@ -50,6 +50,11 @@ generate_alpha_test_single <-
            t.level = NULL,
            group.var,
            adj.vars) {
+
+    if (is.null(alpha.name)){
+      return()
+    }
+
     if (!is.null(time.var) & !is.null(t.level)) {
       condition <- paste(time.var, "== '", t.level, "'", sep = "")
       data.obj <- mStat_subset_data(data.obj, condition = condition)
@@ -85,7 +90,17 @@ generate_alpha_test_single <-
 
       # Run lm and create a coefficient table
       lm.model <- lm(formula, data = merged_df)
-      coef.tab <- broom::tidy(summary(lm.model))
+      summary <- summary(lm.model)
+      coef.tab <- summary$coefficients %>%
+        as.data.frame() %>%
+        rownames_to_column("term") %>%
+        rename(
+          Estimate = "estimate",
+          `Std. Error` = "std.error",
+          `t value` = "statistic",
+          `Pr(>|t|)` = "p.value"
+        ) %>%
+        as_tibble()
 
       # Rearrange the table
       coef.tab <-
@@ -99,14 +114,18 @@ generate_alpha_test_single <-
 
       # Run ANOVA on the model if group.var is multi-categorical
       if (length(na.omit(unique(merged_df[[group.var]]))) > 2) {
-        anova.tab <- broom::tidy(anova(lm.model))
+        anova <- anova(lm.model)
+        anova.tab <- anova %>%
+          as.data.frame() %>%
+          rownames_to_column("term") %>%
+          rename(`F value` = "statistic",
+                 `Pr(>F)` = "p.value")
 
         # Rearrange the table and add missing columns
         anova.tab <- anova.tab %>%
           dplyr::select(
-            term = term,
+            term,
             Statistic = statistic,
-            df = df,
             P.Value = p.value
           ) %>%
           dplyr::mutate(Estimate = NA, Std.Error = NA)
@@ -125,7 +144,7 @@ generate_alpha_test_single <-
           )
 
         coef.tab <-
-          rbind(coef.tab, anova.tab) # Append the anova.tab to the coef.tab
+          rbind(coef.tab, anova.tab)
       }
 
       return(coef.tab)

diff --git a/R/generate_alpha_trend_test_long.R b/R/generate_alpha_trend_test_long.R
@@ -76,6 +76,10 @@ generate_alpha_trend_test_long <- function(data.obj,
                                            group.var = NULL,
                                            adj.vars = NULL) {
 
+  if (is.null(alpha.name)){
+    return()
+  }
+
   if (is.null(alpha.obj)) {
     if (!is_rarefied(data.obj)) {
       message(

diff --git a/R/generate_alpha_volatility_test_long.R b/R/generate_alpha_volatility_test_long.R
@@ -41,7 +41,7 @@
 #' time.var = "visit_number_num",
 #' subject.var = "subject_id",
 #' group.var = "subject_race",
-#' adj.vars = "sample_body_site"
+#' adj.vars = NULL
 #' )
 #' }
 #' @export
@@ -53,6 +53,11 @@ generate_alpha_volatility_test_long <- function(data.obj,
                                                 subject.var,
                                                 group.var,
                                                 adj.vars = NULL) {
+
+  if (is.null(alpha.name)){
+    return()
+  }
+
   if (is.null(alpha.obj)) {
     if (!is_rarefied(data.obj)) {
       message(
@@ -162,14 +167,19 @@ generate_alpha_volatility_test_long <- function(data.obj,
 
     # Run ANOVA on the model if group.var is multi-categorical
     if (length(unique(alpha_df[[group.var]])) > 2) {
-      anova.tab <- broom::tidy(anova(test_result))
+      anova <- anova(test_result)
+      anova.tab <- anova %>%
+        as.data.frame() %>%
+        rownames_to_column("term") %>%
+        rename(`F value` = "statistic",
+               `Pr(>F)` = "p.value") %>%
+        as_tibble()
 
       # Rearrange the table and add missing columns
       anova.tab <- anova.tab %>%
         dplyr::select(
           term = term,
           Statistic = statistic,
-          df = df,
           P.Value = p.value
         ) %>%
         dplyr::mutate(Estimate = NA, Std.Error = NA)

diff --git a/R/generate_beta_change_boxplot_pair.R b/R/generate_beta_change_boxplot_pair.R
@@ -107,6 +107,10 @@ generate_beta_change_boxplot_pair <-
            pdf.hei = 8.5,
            ...) {
 
+    if (is.null(dist.name)){
+      return()
+    }
+
     if (is.null(dist.obj)) {
       meta_tab <- data.obj$meta.dat %>% dplyr::select(all_of(c(subject.var, time.var, group.var, strata.var, adj.vars)))
       dist.obj <-

diff --git a/R/generate_beta_change_spaghettiplot_long.R b/R/generate_beta_change_spaghettiplot_long.R
@@ -94,6 +94,11 @@ generate_beta_change_spaghettiplot_long <-
            ...) {
     # Data validation
     mStat_validate_data(data.obj)
+
+    if (is.null(dist.name)){
+      return()
+    }
+
     if (!is.character(subject.var))
       stop("`subject.var` should be a character string.")
     if (!is.character(time.var))