CODEWITHSUNDEEP
rreplace
Epic_Yarin_God
Epic_Yarin_God

Reputation: 107

Error in `$<-.data.frame ..... replacement has X rows, data has Y

as a part of a project, I was given a medical dataframe for analasys. As part of the analasys, I had to fix the variable called "Gender", due to an error created by reading the data:

datS$Gender <- ifelse(datS$Gender == "¦ֳ«ֳ","Male", ifelse(datS$Gender == "°ֳ·ֳ¡ֳ₪","Female", datS$Gender))

However, doing so, changed the values of the variable to:

 1    2    3 
4461 3373  479

While those numbers are correct, their lable isn't. I attempted to fix that using the replace function:

datS$Gender <- replace(datS$Gender, c("1","2", "3"), c("Male", "Female", "censored"))

However, doing so reasulted in thje following error message:

Error in `$<-.data.frame`(`*tmp*`, Gender2, value = c("3", "3", "3", "3",  : 
  replacement has 8316 rows, data has 8313

The correct number of observasions is 8313, and I don't know what had caused the error. Can you please shed some light over why does this error occure? Many thanks in advance!

Upvotes: 0

Views: 4272

Answers (3)

akrun
akrun

Reputation: 886938

We can convert the factor columns to character with type.convert

dat <- type.convert(dat, as.is = TRUE)

Upvotes: 0

Davide Lorino
Davide Lorino

Reputation: 915

Try a dplyr approach to replacing the strange values:

library(dplyr)

dat %>%
    dplyr::mutate(
        FixedGender = dplyr::case_when(Gender == "¦ֳ«ֳ" ~ "Male",
                                       Gender == "°ֳ·ֳ¡ֳ₪" ~ "Female",
                                       TRUE ~ as.character(Gender))) %>%
    select(Gender, FixedGender) # This line is just to compare the two side-by-side

Upvotes: 1

Ronak Shah
Ronak Shah

Reputation: 388797

Maybe you have columns of factor type. Try turning it to character and then try your ifelse code.

datS$Gender <- as.character(datS$Gender)

You can also have a look at dplyr::recode to recode the values since replace cannot change multiple values.

Upvotes: 1

Related Questions