使用vegan包specaccum计算多采样点物种累积时遇维度错误求助

背景 #

数据集包含120个调查点，每个调查点每年开展4次调查（共480次）。每次15分钟调查按每3分钟时段（共5个时段）记录物种存在/缺失状态，目标是用vegan包的specaccum函数，以collector方法计算每个「调查点+调查日」组合的物种累积（即每新增一个3分钟时段的物种数量变化）。

当前实现代码 #

# Loading packages
library(vegan)

# setwd first
Toronto_data=read.csv(file.choose()) 

# Converting columns that won't be used in the species accumulation curve into factors
cols_to_convert <- c("SurveyDay", "TimeBlock")
Toronto_data[cols_to_convert] <- lapply(Toronto_data[cols_to_convert], as.factor) 

# Split the data frame by multiple factors: Point ID and Number of Days Survey Occurred (~120 points x 4 survey dates)
list_of_samples <- split(Toronto_data, interaction(Toronto_data$PointID, Toronto_data$SurveyDay))

# Remove empty data frames from split_data
list_of_samples <- list_of_samples[sapply(list_of_samples, nrow) > 0]

# Create function to apply specaccum function to numeric columns. Want to use "collector" rather than "exact but it will not work.  
specaccum_function <- function(data) {
  # Ensure that only numeric columns are passed
  numeric_data <- data[sapply(data, is.numeric)]
  
  # Check if numeric_data has more than one row and one column
  if (nrow(numeric_data) > 1 && ncol(numeric_data) > 1) {
    return(specaccum(numeric_data, "collector"))
  } else {
    stop("Not enough data to calculate species accumulation.")
   
  }
}

# Apply specaccum to each subset
results <- lapply(list_of_samples, function(list_of_samples) {
  specaccum_function(list_of_samples)
})

错误信息 #

Error in rowSums(apply(x[ind, , drop = FALSE], 2, cumsum) > 0) : 
  'x' must be an array of at least two dimensions

数据结构示例 #

structure(list(PointID = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L), SurveyDay = c(1L, 
1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 4L, 4L, 
4L, 4L, 4L), TimeBlock = c(1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 
5L, 1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L), SpeciesA = c(1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L), SpeciesB = c(1L, 1L, 1L, 1L, 0L, 0L, 0L, 0L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 1L, 1L), SpeciesC = c(0L, 
0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 
0L, 0L, 0L), SpeciesD = c(1L, 0L, 0L, 0L, 0L, 0L, 1L, 1L, 1L, 
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L), SpeciesE = c(0L, 
0L, 0L, 0L, 0L, 0L, 1L, 1L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 
0L, 0L, 0L), SpeciesF = c(0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L)), class = "data.frame", row.names = c(NA, 
-20L))

问题原因与修正方案 #

错误原因 #

原代码中numeric_data <- data[sapply(data, is.numeric)]会把PointID（整数型，属于numeric）也纳入计算，但specaccum要求输入的是纯物种存在/缺失矩阵（行=调查时段，列=物种）。每个「调查点+调查日」子集中的PointID是固定值，这一列会干扰collector方法的累积计算逻辑，导致内部处理时出现维度异常。

修正代码 #

修改specaccum_function，明确筛选物种列（这里假设物种列以Species开头，可根据实际列名调整）：

specaccum_function <- function(data) {
  # 筛选物种列（仅保留列名以Species开头的列）
  species_data <- data[, grepl("^Species", colnames(data))]
  
  # 检查数据维度：至少2个时段（行）和2个物种（列）
  if (nrow(species_data) > 1 && ncol(species_data) > 1) {
    # 转换为矩阵格式，避免dataframe可能的兼容性问题
    return(specaccum(as.matrix(species_data), method = "collector"))
  } else {
    # 用warning替代stop，避免单个无效子集中断整个批量计算
    warning("Not enough data to calculate species accumulation for this subset.")
    return(NULL)
  }
}

补充说明 #

• 如果物种列的命名规则不是Species开头，可以手动指定列名，比如species_data <- data[, c("SpeciesA", "SpeciesB", "SpeciesC", ...)]
• 转换为矩阵格式是为了确保specaccum处理时的稳定性，避免dataframe的潜在问题

内容的提问来源于stack exchange，提问作者pprest00