Replacing None with 0 in R

Question

So I have data that has None interspersed in numeric vectors, like this -

Lot.Frontage    Lot.Area
34                  3901
70                  8400
60                  7200
64                  7018
111                 16259
50                  4280
155                 20064
60                  7200
70                  9100
None            6449
55                  7642
None            28698

I want to replace the None with 0.

I've tried this

ames.data[ames.data == "None"] <- 0

But this gives me an < NA > wherever there was a none.

How do I replace the None with 0?

`ames.data$Lot.Frontage[ames.data$Lot.Frontage == "None"] <- 0` ? If `ames.data` were a matrix, you would be fine, but I guess it's a data.frame. — Frank, Mar 14 '14 at 20:50
so what I posted above is a subset of the data - I have a lot more columns like that. Doing ames.data[ames.data == "None"] <- 0 has the same issue as in the problem — praks5432, Mar 14 '14 at 20:54
Please provide `dput(ames.data)` or otherwise create reproducible data. http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example — Frank, Mar 14 '14 at 20:58
The `` is because `class(ames.data$Lot.Frontage)` is a `factor` — Jake Burkhead, Mar 14 '14 at 21:01

score 2 · Accepted Answer · answered Mar 14 '14 at 21:07

Here's how to do it over all columns. I generated some "None"s in Lot.Area to show

sapply(ames.data, class)
## Lot.Frontage     Lot.Area                                                                                                                                                                                                                  
##     "factor"    "integer"                                                                                                                                                                                                                  

ames.data$Lot.Area <- ifelse(runif(nrow(ames.data)) < 0.25, "None", ames.data$Lot.Area)
##    Lot.Frontage Lot.Area                                                                                                                                                                                                                   
## 1            34     3901                                                                                                                                                                                                                   
## 2            70     None                                                                                                                                                                                                                   
## 3            60     None                                                                                                                                                                                                                   
## 4            64     7018                                                                                                                                                                                                                   
## 5           111    16259                                                                                                                                                                                                                   
## 6            50     4280                                                                                                                                                                                                                   
## 7           155     None                                                                                                                                                                                                                   
## 8            60     None                                                                                                                                                                                                                   
## 9            70     9100                                                                                                                                                                                                                   
## 10         None     None                                                                                                                                                                                                                   
## 11           55     7642                                                                                                                                                                                                                   
## 12         None    28698                                                                                                                                                                                                                   

ames.data <- as.data.frame(lapply(ames.data, function(x) {
  x <- as.character(x)
  x[x == "None"] <- 0
  as.numeric(x)
}))
##    Lot.Frontage Lot.Area                                                                                                                                                                                                                   
## 1            34     3901                                                                                                                                                                                                                   
## 2            70        0                                                                                                                                                                                                                   
## 3            60        0                                                                                                                                                                                                                   
## 4            64     7018                                                                                                                                                                                                                   
## 5           111    16259                                                                                                                                                                                                                   
## 6            50     4280                                                                                                                                                                                                                   
## 7           155        0                                                                                                                                                                                                                   
## 8            60        0                                                                                                                                                                                                                   
## 9            70     9100                                                                                                                                                                                                                   
## 10            0        0                                                                                                                                                                                                                   
## 11           55     7642                                                                                                                                                                                                                   
## 12            0    28698                                                                                                                                                                                                                   

sapply(ames.data, class)
## Lot.Frontage     Lot.Area                                                                                                                                                                                                                  
##    "numeric"    "numeric"

score 1 · Answer 2 · edited Mar 14 '14 at 21:01

1

Check class(ames.data$Lot.Frontage). I bet it is a factor. That means you can only replace values by other values in levels(ames.data$Lot.Frontage).

You can do this a couple of ways, but they all boil down to converting the columns to a type you can change. In this case, convert to character first, then change "None" to "0", then convert to numeric.

ames.data$Lot.Frontage <- as.character(ames.data$Lot.Frontage)
ames.data$Lot.Frontage[ames.data$Lot.Frontage == "None"] <- "0"
ames.data$Lot.Frontage <- as.numeric(ames.data$Lot.Frontage)

If you convert directly to numeric, the "None"s will become NAs. Since you may have other missing data, the "None"s and the other missing data will get confused.

edited Mar 14 '14 at 21:01

Jake Burkhead

6,435
2
21
32

answered Mar 14 '14 at 21:01

Christopher Louden

7,540
2
26
29

ah that was it - is there a way to do this over all of the columns? – praks5432 Mar 14 '14 at 21:05
`lapply(ames.dat, function(col) { col<-as.character(col); col[col == "None"] <- "0"; col})` ... untested. – IRTFM Mar 15 '14 at 04:37

score 0 · Answer 3 · answered Mar 14 '14 at 21:17

0

If you read the data specifying na.strings="None" and colClasses=c("numeric","numeric") you can replace the "None" with 0 as usual

read.table("file", header=T, quote="\"",colClasses=c("numeric","numeric"),na.strings="None")
df[is.na(df)]<-0

answered Mar 14 '14 at 21:17

DatamineR

10,428
3
25
45

score 0 · Answer 4 · answered Nov 14 '22 at 05:41

Using dplyr, you can generalize this function across all columns that are of character type. This is particularly useful when working with a time series, where you have date column.

library(dplyr)

ames.data <- ames.data %>% 
        mutate(across(where(is.character), ~na_if(., "None"))) %>% 
        mutate(across(where(is.character), as.double))

date	col1	col2
{date}	{chr}	{chr}
2000-01-01	1	2
2000-01-02	4	None
2000-01-03	None	6
2000-01-04	3	0

will become:

date	col1	col2
{date}	{dbl}	{dbl}
2000-01-01	1	2
2000-01-02	4	NA
2000-01-03	NA	6
2000-01-04	3	0

Replacing None with 0 in R

4 Answers4