CODEWITHSUNDEEP
r
user4918087
user4918087

Reputation: 421

Combine 2 columns into 1; removing NAs; Incorrect number of Dimensions in R

I have a 8 column data frame. I want to combine 2 columns of a data frame into one (column 6 and 8.) Both of these columns contain the same data. I basically want to remove the "NAs" such that there are 3 rows of "South West" and last row of "South East"

6th column 8th Column
NA ----------- South West
NA ----------- South West
NA ----------- South West
South East -- NA

I subset the data frame to only pull columns 6 and 8 (the columns that I would like to combine.) region_merge = df[,c(6,8)]

I created my own function as below:

comb_region <- function(df){ if (df[,6]=="NA") df[,6] = df[,8] else df[,6] = df[,6] }

I then used the apply function to repeat the function across all my rows:

apply(X = region_merge, MARGIN = 1, FUN = comb_region)

However, I get the error: Error in df[, 6] : incorrect number of dimensions

I'm new to R. Could someone help me understand why I am getting this error?

Alternatively - Do you guys know an alternative solution to merge 2 columns in 1, by removing the "NAs"?

Thanks in advance!!

Upvotes: 1

Views: 233

Answers (1)

W7GVR
W7GVR

Reputation: 2000

Try:

comb_region <- function(df)
 {
 df[,6]=ifelse(is.na(df[,6]), df[,8], df[,6]);
 df[,8]=ifelse(is.na(df[,8]), df[,6], df[,8]);
 return(df)
 }
filled_df=comb_region(df_with_nas)

Direct comparisons to NA typically do not work: use is.na instead.

And notice that the region_merge data.frame you are passing to comb_region only has two columns.

Upvotes: 1

Related Questions