R Shiny服务器部署报错：执行Remove操作触发因子'max'无意义错误

R Shiny脚本接收data_tran和contam_datafile两个数据表，通过Feature.ID列匹配ID，提供三个操作选项：保留匹配ID、移除匹配ID（设为0）、不执行操作。该逻辑在本地合并式ui/server脚本中完全正常，但部署到Ubuntu LTS服务器后，选择"Remove"选项时应用崩溃，错误提示为Error in Summary.factor: ‘max’ not meaningful for factors，其余功能在本地和服务器端均正常。

原响应式代码 #

data_tran_contam_filt_react = reactive({

  if(input$contam_filter == "Remove"){
    data_tran = data_tran_react()
    contam_datafile = contam_datafile()
    rownames(data_tran) = data_tran$Feature.ID
    # data_tran = data_tran[ ! data_tran$Feature.ID %in% contam_datafile$Feature.ID,]
    data_tran[contam_datafile$Feature.ID,!names(data_tran) %in% c("Consensus.Lineage","Feature.ID","ReprSequence","rowID")] = 0
    rownames(data_tran) = 1:nrow(data_tran)
  }
  if(input$contam_filter == "Analyze"){
    data_tran = data_tran_react()
    contam_datafile = contam_datafile()
    data_tran = data_tran[data_tran$Feature.ID %in% contam_datafile$Feature.ID,]
  }
  
  if(input$contam_filter == "No action"){
    data_tran = data_tran_react()
  }
})

完整错误日志 #

Warning in Ops.factor(i, 0L) : ‘>=’ not meaningful for factors
Warning: Error in Summary.factor: ‘max’ not meaningful for factors
  86: stop
  85: Summary.factor
  84: [<-data.frame
  82: <reactive:data_tran_contam_filt_react> [DELETED]
  66: data_tran_contam_filt_react
  65: <reactive:data_long_react> [DELETED]
  49: data_long_react
  45: observe [DELETED]
  44: <observer>
   1: runApp
Warning: Error in Summary.factor: ‘max’ not meaningful for factors
  107: <Anonymous>
  106: stop
  105: data_long_react
  104: exprFunc [DELETED]
  103: widgetFunc
  102: ::
htmlwidgets
shinyRenderWidget
  101: func
   88: renderFunc
   87: renderFunc
   83: renderFunc
   82: output$proc_main_alt
    1: runApp
Warning: Error in Summary.factor: ‘max’ not meaningful for factors
  107: <Anonymous>
  106: stop
  105: data_tran_contam_filt_react
  104: exprFunc [DELETED]
  103: widgetFunc
  102: ::
htmlwidgets
shinyRenderWidget
  101: func
   88: renderFunc
   87: renderFunc
   83: renderFunc
   82: output$proc_new_data
    1: runApp

Execution halted

解决方案 #

问题根源 #

服务器端环境中，data_tran$Feature.ID的类型为因子(factor)，而本地环境中该列是字符型(character)。当用因子作为行名后，后续用contam_datafile$Feature.ID（同样是因子）去索引行时，R会尝试将因子与数值进行比较，导致因子操作相关的错误。

修复代码 #

data_tran_contam_filt_react = reactive({
  # 统一获取基础数据并转换Feature.ID为字符型
  data_tran = data_tran_react()
  data_tran$Feature.ID = as.character(data_tran$Feature.ID)
  
  contam_datafile = contam_datafile()
  contam_datafile$Feature.ID = as.character(contam_datafile$Feature.ID)
  
  if(input$contam_filter == "Remove"){
    # 找到需要设置为0的行索引
    contam_rows = data_tran$Feature.ID %in% contam_datafile$Feature.ID
    # 指定要修改的列（排除非数值列）
    target_cols = !names(data_tran) %in% c("Consensus.Lineage","Feature.ID","ReprSequence","rowID")
    # 赋值操作
    data_tran[contam_rows, target_cols] = 0
  } else if(input$contam_filter == "Analyze"){
    data_tran = data_tran[data_tran$Feature.ID %in% contam_datafile$Feature.ID,]
  }
  # No action分支直接返回原始data_tran，无需额外处理
  
  return(data_tran)
})

关键优化点 #

• 提前将两个数据表的Feature.ID转为字符型，避免因子类型带来的索引错误
• 用逻辑向量contam_rows定位目标行，替代直接用行名索引的方式，更稳定可靠
• 合并数据获取逻辑，避免重复调用data_tran_react()和contam_datafile()
• 使用else if结构简化条件判断逻辑

内容的提问来源于stack exchange，提问作者Alex Umbach