CODEWITHSUNDEEP
rdataframereplace
Manolo Dominguez Becerra
Manolo Dominguez Becerra

Reputation: 1373

Problems replacing character values in a dataframe in R

One of my column in my datafra looks like this:

[1] "MIT"  "MIT"  "NUC"  "MIT"  "CYT"  "MIT"  "NUC"  "MIT"  "CYT"  "NUC"  "NUC"  "CYT"  "MIT"  "MIT"  ""     ""     "CYT" 

[18] "NUC"  "NUC"  "NUC"  "CYT"  "CYT"  "CYT"  "CYT"  "CYT"  "MIT"  "CYT"  "CYT"  "MIT"  "ME1"  "NUC"  "MIT"  "CYT"  "NUC" 

[35] "ME1"  "EXC"  "MIT"  "CYT"  "MIT"  "MIT"  "ME2"  "ME2"  ""     ""     "ME3"  "VAC"  "NUC"  "ME2"  "EXC"  "VAC"  "ME3"

I need to change "ME3" for ME.

When I used this code:

df_norm_pre$MIT["ME1"]="ME"

I find this error:

Error in $<-.data.frame(*tmp*, MIT, value = c("MIT", "MIT", "NUC", : replacement has 1550 rows, data has 1549.

Could someone help me with this problems?

Upvotes: 0

Views: 92

Answers (2)

akrun
akrun

Reputation: 887881

We can use replace 

df_norm_pre$MIT <- replace(df_norm_pre$MIT, df_norm_pre$MIT == "ME3", "ME")

Or just assignment

df_norm_pre$MIT[df_norm_pre$MIT == "ME3"] <- "ME"

Or with sub

df_norm_pre$MIT <-  sub("^ME\\d+$", "ME", df_norm_pre$MIT)

Error in OP's code is because the indexing is not correct

df_norm_pre$MIT["ME1"]

The df_norm_pre$MIT extracts the column as a vector, then using ["ME1"], expects that is either a column with name of "ME1" or a list with that name. Both of these are not TRUE here. Instead, we can use logical index to subset df_norm_pre$MIT == "ME1"

Upvotes: 1

ThomasIsCoding
ThomasIsCoding

Reputation: 102700

Maybe you can try gsub to see if it can do the replacement

df_norm_pre <- within(df_norm_pre, MIT <- gsub("ME3","ME",MIT))

or use replace like

df_norm_pre <- within(df_norm_pre, MIT <- replace(MIT, MIT == "ME3", "ME"))

Upvotes: 1

Related Questions