Address #141: Added plotly

Author: boxuancui

DESCRIPTION

@@ -25,7 +25,8 @@ Suggests:
     covr,
     knitr,
     jsonlite,
-    nycflights13
+    nycflights13,
+    plotly
 SystemRequirements: pandoc (>= 1.12.3) - http://pandoc.org
 License: MIT + file LICENSE
 Language: en-US

NEWS.md

@@ -1,4 +1,7 @@
 # DataExplorer 0.9.0
+## New Features
+* [#141](https://github.com/boxuancui/DataExplorer/issues/141): Added `plotly` argument to all plot functions. When `plotly = TRUE`, plots are converted to interactive plotly objects via `plotly::ggplotly()` (requires the plotly package).
+
 ## Enhancements
 * [#181](https://github.com/boxuancui/DataExplorer/issues/181): Added `by` argument to `plot_histogram` and `plot_density` to break down distributions by a discrete or continuous feature.
 

R/create_report.R

@@ -6,6 +6,7 @@
 #' @param output_file output file name in \link[rmarkdown]{render}. Default is "report.html".
 #' @param output_dir output directory for report in \link[rmarkdown]{render}. Default is user's current directory.
 #' @param y name of response variable if any. Response variables will be passed to appropriate plotting functions automatically.
+#' @param plotly if \code{TRUE}, use interactive plotly charts in the report (requires the \pkg{plotly} package). Default is \code{FALSE}. Only applies to HTML output; PDF reports use static plots.
 #' @param config report configuration generated by \link{configure_report}.
 #' @param report_title report title. Default is "Data Profiling Report".
 #' @param \dots other arguments to be passed to \link[rmarkdown]{render}.
@@ -33,6 +34,13 @@
 #' # Create report
 #' create_report(iris)
 #' create_report(airquality, y = "Ozone")
+#' 
+#' # Create report with plotly
+#' # Note: It is a known issue that some facet panels may not show up in plotly.
+#' # More details in the following issues:
+#' # * https://github.com/plotly/plotly.R/issues/1243
+#' # * https://github.com/plotly/plotly.R/issues/1962
+#' create_report(airquality, y = "Ozone", plotly = TRUE)
 #'
 #' # Load library
 #' library(ggplot2)
@@ -92,6 +100,7 @@ create_report <- function(data,
                           output_file = "report.html",
                           output_dir = getwd(),
                           y = NULL,
+                          plotly = FALSE,
                           config = configure_report(),
                           report_title = "Data Profiling Report",
                           ...) {
@@ -117,7 +126,7 @@ create_report <- function(data,
     output_file = output_file,
     output_dir = output_dir,
     intermediates_dir = output_dir,
-    params = list(data = data, report_config = config, response = y, set_title = report_title),
+    params = list(data = data, report_config = config, response = y, set_title = report_title, plotly = plotly),
     ...
   ))
   ## Open report (use path returned by render in case extension was normalized)

R/plot.R

@@ -5,6 +5,7 @@
 #' @param title plot title
 #' @param ggtheme complete ggplot2 themes
 #' @param theme_config a list of configurations to be passed to \link[ggplot2]{theme}
+#' @param plotly if \code{TRUE}, convert ggplot to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @param \dots other arguments to be passed
 #' @return invisibly return the named list of ggplot objects
 #' @keywords internal
@@ -26,7 +27,7 @@
 #'   "aspect.ratio" = 1
 #'   )
 #' )
-plotDataExplorer <- function(plot_obj, title, ggtheme, theme_config, ...) {
+plotDataExplorer <- function(plot_obj, title, ggtheme, theme_config, plotly = FALSE, ...) {
   UseMethod("plotDataExplorer")
 }
 
@@ -48,20 +49,43 @@ plotDataExplorer <- function(plot_obj, title, ggtheme, theme_config, ...) {
 #' @import gridExtra
 #' @export
 #' @seealso \link{plotDataExplorer} \link{plotDataExplorer.single} \link{plotDataExplorer.multiple}
-plotDataExplorer.grid <- function(plot_obj, title, ggtheme, theme_config, page_layout, nrow, ncol, ...) {
+plotDataExplorer.grid <- function(plot_obj, title, ggtheme, theme_config, page_layout, nrow, ncol, plotly = FALSE, ...) {
   plot_list <- lapply(plot_obj, function(p) {
     p +
       eval(ggtheme) +
       do.call(theme, theme_config)
   })
 
-  if (length(page_layout) > 1) {
-    invisible(lapply(names(page_layout), function(pg_name) {
-      index <- page_layout[[pg_name]]
-      suppressMessages(do.call(grid.arrange, c(plot_list[index], ncol = ncol, nrow = nrow, top = title, bottom = pg_name)))
-    }))
+  if (plotly) {
+    if (!requireNamespace("plotly", quietly = TRUE)) {
+      stop("Package \"plotly\" is required for plotly = TRUE. Install it with install.packages(\"plotly\").")
+    }
+    if (length(page_layout) > 1) {
+      pl_list <- lapply(names(page_layout), function(pg_name) {
+        index <- page_layout[[pg_name]]
+        plist <- lapply(plot_list[index], function(p) suppressWarnings(plotly::ggplotly(p)))
+        plotly::subplot(plist, nrows = nrow, ncols = ncol, margin = 0.05, titleY = TRUE, titleX = TRUE)
+      })
+    } else {
+      plist <- lapply(plot_list, function(p) suppressWarnings(plotly::ggplotly(p)))
+      pl_list <- list(plotly::subplot(plist, nrows = nrow, ncols = ncol, margin = 0.05, titleY = TRUE, titleX = TRUE))
+    }
+    if (isTRUE(getOption("knitr.in.progress"))) {
+      if (requireNamespace("htmltools", quietly = TRUE)) {
+        return(do.call(htmltools::tagList, pl_list))
+      }
+      return(pl_list)
+    }
+    invisible(lapply(pl_list, print))
   } else {
-    suppressMessages(do.call(grid.arrange, c(plot_list, ncol = ncol, nrow = nrow, top = title)))
+    if (length(page_layout) > 1) {
+      invisible(lapply(names(page_layout), function(pg_name) {
+        index <- page_layout[[pg_name]]
+        suppressMessages(do.call(grid.arrange, c(plot_list[index], ncol = ncol, nrow = nrow, top = title, bottom = pg_name)))
+      }))
+    } else {
+      suppressMessages(do.call(grid.arrange, c(plot_list, ncol = ncol, nrow = nrow, top = title)))
+    }
   }
 
   invisible(plot_list)
@@ -81,13 +105,22 @@ plotDataExplorer.grid <- function(plot_obj, title, ggtheme, theme_config, page_l
 #' @import ggplot2
 #' @export
 #' @seealso \link{plotDataExplorer} \link{plotDataExplorer.grid} \link{plotDataExplorer.multiple}
-plotDataExplorer.single <- function(plot_obj, title, ggtheme, theme_config, ...) {
+plotDataExplorer.single <- function(plot_obj, title, ggtheme, theme_config, plotly = FALSE, ...) {
   plot_obj <- plot_obj +
     ggtitle(title) +
     eval(ggtheme) +
     do.call(theme, theme_config)
 
-  print(plot_obj)
+  if (plotly) {
+    if (!requireNamespace("plotly", quietly = TRUE)) {
+      stop("Package \"plotly\" is required for plotly = TRUE. Install it with install.packages(\"plotly\").")
+    }
+    pl <- suppressWarnings(plotly::ggplotly(plot_obj))
+    if (isTRUE(getOption("knitr.in.progress"))) return(pl)
+    print(pl)
+  } else {
+    print(plot_obj)
+  }
   invisible(plot_obj)
 }
 
@@ -108,7 +141,7 @@ plotDataExplorer.single <- function(plot_obj, title, ggtheme, theme_config, ...)
 #' @importFrom stats setNames
 #' @export
 #' @seealso \link{plotDataExplorer} \link{plotDataExplorer.grid} \link{plotDataExplorer.single}
-plotDataExplorer.multiple <- function(plot_obj, title, ggtheme, theme_config, page_layout, facet_wrap_args = list(), ...) {
+plotDataExplorer.multiple <- function(plot_obj, title, ggtheme, theme_config, page_layout, facet_wrap_args = list(), plotly = FALSE, ...) {
   n <- length(page_layout)
   plot_list <- lapply(setNames(seq.int(n), paste0("page_", seq.int(n))), function(i) {
     plot_obj[[i]] +
@@ -119,6 +152,21 @@ plotDataExplorer.multiple <- function(plot_obj, title, ggtheme, theme_config, pa
       do.call(theme, theme_config)
   })
 
-  invisible(capture.output(print(plot_list)))
+  if (plotly) {
+    if (!requireNamespace("plotly", quietly = TRUE)) {
+      stop("Package \"plotly\" is required for plotly = TRUE. Install it with install.packages(\"plotly\").")
+    }
+    pl_list <- lapply(plot_list, function(p) suppressWarnings(plotly::ggplotly(p)))
+    if (isTRUE(getOption("knitr.in.progress"))) {
+      if (requireNamespace("htmltools", quietly = TRUE)) {
+        return(do.call(htmltools::tagList, pl_list))
+      }
+      lapply(pl_list, print)
+      return(invisible(plot_list))
+    }
+    invisible(lapply(pl_list, print))
+  } else {
+    invisible(capture.output(print(plot_list)))
+  }
   invisible(plot_list)
 }

R/plot_bar.R

@@ -14,6 +14,7 @@
 #' @param nrow number of rows per page. Default is 3.
 #' @param ncol number of columns per page. Default is 3.
 #' @param parallel enable parallel? Default is \code{FALSE}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the named list of ggplot objects
 #' @keywords plot_bar
 #' @details If a discrete feature contains more categories than \code{maxcat} specifies, it will not be passed to the plotting function.
@@ -34,14 +35,17 @@
 #' # Plot bar charts by `cut`
 #' plot_bar(diamonds, by = "cut")
 #' plot_bar(diamonds, by = "cut", by_position = "dodge")
+#'
+#' # Interactive plotly version (requires the plotly package)
+#' # plot_bar(diamonds, plotly = TRUE)
 
 plot_bar <- function(data, with = NULL,
                      by = NULL, by_position = "fill",
                      maxcat = 50, order_bar = TRUE, binary_as_factor = TRUE,
                      title = NULL,
                      ggtheme = theme_gray(), theme_config = list(),
                      nrow = 3L, ncol = 3L,
-                     parallel = FALSE) {
+                     parallel = FALSE, plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   frequency <- measure <- variable <- value <- facet_value <- NULL
   ## Check if input is data.table
@@ -120,6 +124,7 @@ plot_bar <- function(data, with = NULL,
     title = title,
     ggtheme = ggtheme,
     theme_config = theme_config,
+    plotly = plotly,
     facet_wrap_args = list(
       "facet" = ~ variable,
       "nrow" = nrow,

R/plot_boxplot.R

@@ -13,6 +13,7 @@
 #' @param nrow number of rows per page
 #' @param ncol number of columns per page
 #' @param parallel enable parallel? Default is \code{FALSE}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the named list of ggplot objects
 #' @keywords plot_boxplot
 #' @import data.table
@@ -41,7 +42,7 @@ plot_boxplot <- function(data, by,
                          title = NULL,
                          ggtheme = theme_gray(), theme_config = list(),
                          nrow = 3L, ncol = 4L,
-                         parallel = FALSE) {
+                         parallel = FALSE, plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   variable <- by_f <- value <- NULL
   ## Check if input is data.table
@@ -85,6 +86,7 @@ plot_boxplot <- function(data, by,
     title = title,
     ggtheme = ggtheme,
     theme_config = theme_config,
+    plotly = plotly,
     facet_wrap_args = list(
       "facet" = ~ variable,
       "nrow" = nrow,

R/plot_correlation.R

@@ -9,6 +9,7 @@
 #' @param title plot title
 #' @param ggtheme complete ggplot2 themes. The default is \link[ggplot2]{theme_gray}.
 #' @param theme_config a list of configurations to be passed to \link[ggplot2]{theme}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the ggplot object
 #' @keywords plot_correlation
 #' @details For discrete features, the function first dummifies all categories, then calculates the correlation matrix (see \link[stats]{cor}) and plots it.
@@ -27,7 +28,8 @@ plot_correlation <- function(data, type = c("all", "discrete", "continuous"), ma
                              title = NULL,
                              ggtheme = theme_gray(),
                              theme_config = list("legend.position" = "bottom",
-                                                 "axis.text.x" = element_text(angle = 90))) {
+                                                 "axis.text.x" = element_text(angle = 90)),
+                             plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   Var1 <- Var2 <- value <- NULL
   ## Set data to data.table
@@ -73,6 +75,7 @@ plot_correlation <- function(data, type = c("all", "discrete", "continuous"), ma
     plot_obj = output,
     title = title,
     ggtheme = ggtheme,
-    theme_config = theme_config
+    theme_config = theme_config,
+    plotly = plotly
   )
 }

R/plot_density.R

@@ -12,6 +12,7 @@
 #' @param nrow number of rows per page. Default is 4.
 #' @param ncol number of columns per page. Default is 4.
 #' @param parallel enable parallel? Default is \code{FALSE}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the named list of ggplot objects
 #' @keywords plot_density
 #' @import data.table
@@ -40,7 +41,7 @@ plot_density <- function(data, by = NULL, binary_as_factor = TRUE,
                          title = NULL,
                          ggtheme = theme_gray(), theme_config = list(),
                          nrow = 4L, ncol = 4L,
-                         parallel = FALSE) {
+                         parallel = FALSE, plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   variable <- value <- by_f <- NULL
   ## Check if input is data.table
@@ -95,6 +96,7 @@ plot_density <- function(data, by = NULL, binary_as_factor = TRUE,
     title = title,
     ggtheme = ggtheme,
     theme_config = theme_config,
+    plotly = plotly,
     facet_wrap_args = list(
       "facet" = ~ variable,
       "nrow" = nrow,

R/plot_histogram.R

@@ -12,6 +12,7 @@
 #' @param nrow number of rows per page. Default is 4.
 #' @param ncol number of columns per page. Default is 4.
 #' @param parallel enable parallel? Default is \code{FALSE}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the named list of ggplot objects
 #' @keywords plot_histogram
 #' @import data.table
@@ -37,7 +38,7 @@ plot_histogram <- function(data, by = NULL, binary_as_factor = TRUE,
                            title = NULL,
                            ggtheme = theme_gray(), theme_config = list(),
                            nrow = 4L, ncol = 4L,
-                           parallel = FALSE) {
+                           parallel = FALSE, plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   variable <- value <- by_f <- NULL
   ## Check if input is data.table
@@ -92,6 +93,7 @@ plot_histogram <- function(data, by = NULL, binary_as_factor = TRUE,
     title = title,
     ggtheme = ggtheme,
     theme_config = theme_config,
+    plotly = plotly,
     facet_wrap_args = list(
       "facet" = ~ variable,
       "nrow" = nrow,

R/plot_intro.R

@@ -6,6 +6,7 @@
 #' @param title plot title
 #' @param ggtheme complete ggplot2 themes. The default is \link[ggplot2]{theme_gray}.
 #' @param theme_config a list of configurations to be passed to \link[ggplot2]{theme}.
+#' @param plotly if \code{TRUE}, convert to interactive plotly object (requires the \pkg{plotly} package). Default is \code{FALSE}.
 #' @return invisibly return the ggplot object
 #' @keywords plot_intro
 #' @import ggplot2
@@ -17,7 +18,7 @@
 #' plot_intro(airquality)
 #' plot_intro(iris)
 
-plot_intro <- function(data, geom_label_args = list(), title = NULL, ggtheme = theme_gray(), theme_config = list()) {
+plot_intro <- function(data, geom_label_args = list(), title = NULL, ggtheme = theme_gray(), theme_config = list(), plotly = FALSE) {
   ## Declare variable first to pass R CMD check
   id <- dimension <- variable <- value <- NULL
   ## Get intro data
@@ -54,6 +55,7 @@ plot_intro <- function(data, geom_label_args = list(), title = NULL, ggtheme = t
     plot_obj = output,
     title = ifelse(is.null(title), paste("Memory Usage:", memory_usage_string), title),
     ggtheme = ggtheme,
-    theme_config = theme_config
+    theme_config = theme_config,
+    plotly = plotly
   )
 }