使用 grepl 的 for 循环

Question

I am trying to use a for loop to run through a data frame, see if an observation contains a certain string in a column (ie, it should contain "no law" in the column Content ), and generate values in a different column based on the outcome.

If it does contain the string, which is identified by the outcome of the grepl function being True , then the observation should have 'Permissive' in the Effectrp column; otherwise, it should say 'Restrictive'.

I'm not quite sure what I'm doing wrong... Any help would be appreciated!

for (i in 1:nrow(ldb)){
  if (grepl('no law', ldb$Content[i], ignore.case = TRUE)) == TRUE {
    ldb$Effectrp[i] = 'Permissive'
  } else {
    lab$EffectTR[i] = 'Restrictive'
  }
}

Answer 1

You shouldn't even need a for loop for this, as grepl will return a vector if applied to a vector. You could try something like

ldb$Effectrp <- 'Restrictive'
lbd$Effectrp[grepl('no law', ldb$Content, ignore.case = TRUE)] <- 'Permissive'

(and, as mentioned in the previous answer, be careful about the typos in your data frame and column names.)

Answer 2

Using base R :

ldb$EffectRP <- sapply(ldb$Content, 
                       function(x) if (grepl("no law", x, ignore.case = TRUE)) {"Permissive"} else {"Restrictive"} )

Using dplyr and stringr :

ldb %>%
  mutate(EffectRP2 = ifelse(str_detect(Content, "no law"), "Permissive", "Restrictive"))

Those options return:

   Content    EffectRP   EffectRP2
1      law Restrictive Restrictive
2   no law  Permissive  Permissive
3      law Restrictive Restrictive
4   no law  Permissive  Permissive
5      law Restrictive Restrictive
6      law Restrictive Restrictive
7   no law  Permissive  Permissive
8   no law  Permissive  Permissive
9   no law  Permissive  Permissive
10  no law  Permissive  Permissive

Answer 3

Similar to the answer I wrote to this question.

The only potential problem I can see is typos in the else part. You wrote lab$EffectTR when earlier your dataframe was named ldb and the column was named Effectrp . Not sure if this is intentional.

Clarification of the redundancy thing:

You don't need the == TRUE in your if statement. I think of it like this. Right now you have:

if (grepl(check if my pattern is found in Content) == TRUE) {
    do something
}

grepl returns TRUE/FALSE , so let's say "no law" is found in Content , then grepl evaluates to TRUE , producing:

if (TRUE == TRUE) {
    do something
}

If we continue evaluating the parentheses, we know that indeed, TRUE == TRUE , so this reduces to:

if (TRUE) {
    do something
}

This is what we want. However, the extra check TRUE == TRUE is unnecessary when you could just use the output from grepl like so:

if (grepl(check if my pattern is found in Content) {
    do something
}

This will evaluate to:

if (TRUE) {
    do something
}

the same thing as before, but you're skipping the redundant == TRUE step. ie the if statement will run, because the thing inside the parentheses is TRUE in a quite literal sense.

Hope that makes more sense. It was confusing to me when I first learned it as well.