R语言networkD3包radialNetwork遇NA值报错，需跳过节点方案

Hey there! The NA values in your dataset are breaking the hierarchical structure that data.tree and radialNetwork rely on—let's fix that with two straightforward approaches tailored to your workflow.

Approach 1: Preprocess Data to Exclude NA Levels Before Building the Tree #

This method cleans up your path strings upfront, so you never introduce NA nodes into the tree in the first place. We'll filter out NA values from each row's hierarchy before constructing the pathString.

# Load required packages
library(data.tree)
library(networkD3)

# Read your custom data (replace with your file path)
your_df <- read.csv("your_custom_data.csv", stringsAsFactors = FALSE)

# Build pathString by skipping NA values in each row's hierarchy
your_df$pathString <- apply(your_df[, c("session", "room", "speaker")], 1, function(x) {
  # Start with the root node, then add only non-NA child levels
  valid_hierarchy <- c("useR", x[!is.na(x)])
  paste(valid_hierarchy, collapse = "|")
})

# Optional: Filter out rows where all child levels are NA (avoids empty root child nodes)
your_df <- your_df[apply(your_df[, c("session", "room", "speaker")], 1, function(x) any(!is.na(x))), ]

# Build the tree and render the radial network
your_tree <- as.Node(your_df, pathDelimiter = "|")
your_tree_list <- ToListExplicit(your_tree, unname = TRUE)
radialNetwork(your_tree_list)

Approach 2: Prune NA Nodes After Building the Tree #

If you prefer to keep your original pathString construction logic, you can build the full tree first, then remove any nodes with NA values using data.tree's Prune function.

# Load required packages
library(data.tree)
library(networkD3)

# Read your custom data
your_df <- read.csv("your_custom_data.csv", stringsAsFactors = FALSE)

# Build pathString as you originally did (including NA values)
your_df$pathString <- paste("useR", your_df$session, your_df$room, your_df$speaker, sep="|")

# Construct the tree
your_tree <- as.Node(your_df, pathDelimiter = "|")

# Prune all nodes with NA as their name
Prune(your_tree, function(node) !is.na(node$name))

# Convert to list and render the radial network
your_tree_list <- ToListExplicit(your_tree, unname = TRUE)
radialNetwork(your_tree_list)

Quick Note #

Both methods work to eliminate NA-related errors, but Approach 1 is more efficient (it avoids creating invalid nodes altogether) while Approach 2 gives you visibility into the original tree structure before cleanup. Pick whichever fits your debugging and data needs best!

内容的提问来源于stack exchange，提问作者L.Ferg