当我在循环中使用write.xlsx时,如何处理空数据帧?
下面是循环的样子,其中source(" ./ Scripts / Analysis_details.R")引用了创建数据框的r文件。
library(xlsx)
for (A in unique(df_base$A)) {
df<- df_base[df_base$A==A,]
source("./Scripts/Analysis_details.R")
output_file = paste("./Output/report_", A, '_', Sys.Date(), ".xlsx", sep='')
write.xlsx(df1, file=output_file, sheetName="df1", append=TRUE, row.names=FALSE, showNA = FALSE)
write.xlsx(df2, file=output_file, sheetName="df2", append=TRUE, row.names=FALSE, showNA = FALSE)
write.xlsx(df3, file=output_file, sheetName="df3", append=TRUE, row.names=FALSE, showNA = FALSE)
write.xlsx(df4, file=output_file, sheetName="df4", append=TRUE, row.names=FALSE, showNA = FALSE)}
我得到的错误是......
Error in mapply(setCellValue, cells[seq_len(nrow(cells)), colIndex[ic]], : zero-length inputs cannot be mixed with those of non-zero length
答案 0 :(得分:3)
我能够通过将修改后的write.xlsx()
函数放入我的脚本来解决这个问题。如果数据框的行为零,则跳过.write_block()
,并且仅使用列名保存文件。请注意,您还必须将原始.write_block()
函数复制到脚本中。
write.xlsx.custom <- function(x, file, sheetName="Sheet1",
col.names=TRUE, row.names=TRUE, append=FALSE, showNA=TRUE)
{
if (!is.data.frame(x))
x <- data.frame(x) # just because the error message is too ugly
iOffset <- jOffset <- 0
if (col.names)
iOffset <- 1
if (row.names)
jOffset <- 1
if (append && file.exists(file)){
wb <- loadWorkbook(file)
} else {
ext <- gsub(".*\\.(.*)$", "\\1", basename(file))
wb <- createWorkbook(type=ext)
}
sheet <- createSheet(wb, sheetName)
noRows <- nrow(x) + iOffset
noCols <- ncol(x) + jOffset
if (col.names){
rows <- createRow(sheet, 1) # create top row
cells <- createCell(rows, colIndex=1:noCols) # create cells
mapply(setCellValue, cells[1,(1+jOffset):noCols], colnames(x))
}
if (row.names) # add rownames to data x
x <- cbind(rownames=rownames(x), x)
if(nrow(x) > 0) {
colIndex <- seq_len(ncol(x))
rowIndex <- seq_len(nrow(x)) + iOffset
.write_block(wb, sheet, x, rowIndex, colIndex, showNA)
}
saveWorkbook(wb, file)
invisible()
}
找到了原始功能