在 R 中创建循环以清理多个数据集

Question

Suppose I have the following dataframes (essentially 4 similar datasets):

q1_hosp <- data.frame(dxe1 = c(1,NULL, NA, NULL, 1), dxe2 = c(1,NULL, NULL, NULL, 1))
q2_hosp <- data.frame(dxe1 = c(NULL,1, 1, NA, 1), dxe2 = c(1,1, 1, NULL, 1))
q3_hosp <- data.frame(dxe1 = c(NA,NULL, 1 1, 1), dxe2 = c(1,NA, NA, NULL, 1))
q4_hosp <- data.frame(dxe1 = c(1,1, 1 NULL, 1), dxe2 = c(1,NULL, 1, 1, 1))

What I would like is to turn the NULL entries into missing values in R. I can do them by writing the following codes in a tedious manner:

any(is.na(q1_hosp$dxe1)) # there are missing rows, keep NAs
q1_hosp$dxe1 <- na_if(q1_hosp$dxe1, "NULL") #convert NULLto NA
any(is.na(q1_hosp$dxe2)) # there are missing rows, keep NAs
q1_hosp$dxe2 <- na_if(q1_hosp$dxe2, "NULL") #convert NULLto NA

any(is.na(q2_hosp$dxe1)) # there are missing rows, keep NAs
q2_hosp$dxe1 <- na_if(q2_hosp$dxe1, "NULL") #convert NULLto NA
any(is.na(q2_hosp$dxe2)) # there are missing rows, keep NAs
q2_hosp$dxe2 <- na_if(q2_hosp$dxe2, "NULL") #convert NULLto NA

any(is.na(q3_hosp$dxe1)) # there are missing rows, keep NAs
q3_hosp$dxe1 <- na_if(q3_hosp$dxe1, "NULL") #convert NULLto NA
any(is.na(q3_hosp$dxe2)) # there are missing rows, keep NAs
q3_hosp$dxe2 <- na_if(q3_hosp$dxe2, "NULL") #convert NULLto NA

any(is.na(q4_hosp$dxe1)) # there are missing rows, keep NAs
q4_hosp$dxe1 <- na_if(q4_hosp$dxe1, "NULL") #convert NULLto NA
any(is.na(q4_hosp$dxe2)) # there are missing rows, keep NAs
q4_hosp$dxe2 <- na_if(q4_hosp$dxe2, "NULL") #convert NULLto NA

What I would like is to create a loop so that the process is less tedious.

I have been trying with this

  for(i in 1:4) {
  ##Clean the dxe variables
  any(is.na(q[i]_hosp$dxe1)) # there are missing rows, keep NAs
  q[i]_hosp$dxe1 <- na_if(q[i]_hosp$dxe1, "NULL") #convert NULLto NA
 any(is.na(q[i]_hosp$dxe2)) # there are missing rows, keep NAs
  q[i]_hosp$dxe1 <- na_if(q[i]_hosp$dxe2, "NULL") #convert NULLto NA
}

But my code doesn't work and I am getting errors. Can anyone help write the for loop for what I want to achieve?

Answer 1

Get the data in a list and use lapply to change 'NULL' to NA from each of the dataframe.

result <- lapply(mget(sprintf('q%d_hosp', 1:4)), function(x) {x[x == 'NULL'] <- NA;x})

If you want the changes in individual dataframes back.

list2env(result, .GlobalEnv)

Answer 2

Try this.

1. all names of dataset are stored in dl , which can access with get function
2. RMNULL function is used to replace NULL by NA
3. lapply function is used to loop all datasets

# dataset name list 
dl <- sprintf("%s%s%s","q",1:2,"_hosp")
df.list <- lapply(dl,get)

# replace NULL with NA
## remove NULL function
RMNULL <- function(df){
    is.na(df) <- df == "NULL"
    return(df)
}
lapply(df.list, RMNULL)

example data:

q1_hosp <- structure(list(x = 1:3, y = list(NULL, NULL, NULL)), .Names = c("x",  "y"), row.names = c(NA, -3L), class = "data.frame")

q2_hosp <- structure(list(x = 4:6, y = list(1, NULL, NULL)), .Names = c("x",  "y"), row.names = c(NA, -3L), class = "data.frame")