setting different conditions on multiple columns of a data frame in one go

I need to write a condition saying if 9 columns of a big data frame include either of the followings: "6%t" (as the first three characters in each row), or NA, or "", then save the file as csv. I have difficulty setting the conditions in the correct way. The aim is to make sure, I save the write data frame.

Assume that cols is the 9 columns that I need to check in my data frame.

cols -> c("AA", "BB", "SS", "EE"", "OO", "UU", "PP", "QQ", "FF")
if (substring(df[,cols], 1, 3) == "6%t" || is.na(df[,cols]) || df[,cols] == "") {
  write.csv(df, file = paste0(path, ".csv"))}

However, I get the following error:

the condition has length > 1

Could you please help me figure this out?

Solution

Always provide some data with your code. The problem is simple: conditions with the wrong dimensions. The IF statement must be TRUE or FALSE, not a logical vector. Look:

#
aux <- sample(c(words, 1000:1999), 17)
aux <- sample(c(aux, "6%tFoo", NA, ""))
aux <- structure(aux, dim = 4:5)
colnames(aux) <- letters[1:5]

     a          b            c          d          e     
[1,] "the"      "particular" "consider" "6%tFoo"   "1649"
[2,] "1167"     "1402"       "of"       "along"    NA    
[3,] "question" ""           "many"     "exercise" "1709"
[4,] "1397"     "1152"       "oppose"   "problem"  "1114"

# 
substring(aux, 1, 3) == "6%t"

     a     b     c     d     e
[1,] FALSE FALSE FALSE  TRUE FALSE
[2,] FALSE FALSE FALSE FALSE    NA
[3,] FALSE FALSE FALSE FALSE FALSE
[4,] FALSE FALSE FALSE FALSE FALSE

Your dataset appears to have NA values, so be aware of substring behavior with it. Try this:

#
test <- \(x) any((sapply(x, substring, 1, 3) == "6%t") & !is.na(x)) ||
  any(is.na(x)) ||
  any(x == "")

#
cols <- "b"
test(aux[, cols])
[1] TRUE

#
cols <- "d"
test(aux[, cols])
[1] TRUE

#
cols <- "e"
test(aux[, cols])
[1] TRUE

#
cols <- c("a", "c")
test(aux[, cols])
[1] FALSE

EDITED after the first reply.

To check if all elements of selected columns are either NA, "" or starts with "6%t":

aux <- structure(
  list(
    a = c("the", "1167", "question", "1397"), 
    b = c("", "", "", NA), 
    c = c("consider", "of", "many", "oppose"), 
    d = c("6%tFoo", "6%talong", "", NA), 
    e = c("1649", NA, "1709", "1114")), 
  
  row.names = c(NA, -4L), 
  class = "data.frame")

param   <- c(NA, "", "6%t")

# Condition must be true
cols    <- c("b", "d") 
aux_sub <- sapply(aux[, cols], substring, 1, 3)

test <- length(setdiff(aux_sub, param)) == 0
[1] TRUE

# Condition must be false
cols    <- c("a", "b", "d")
aux_sub <- sapply(aux[, cols], substring, 1, 3)

test <- length(setdiff(aux_sub, param)) == 0
[1] FALSE