How to Convert a Character to Numeric in R

Basic R Syntax:

x_num <- as.numeric(x)

Do you need more explanations? In the following article, I’ll provide you with all important information for the conversion of character vectors to numeric in R.

Example: Convert Character to Numeric in R

Before we can dive into the transformation of a character variable to numeric, we need to create an example character in R. Consider the following vector:

set.seed(55555)                                              # Set seed
x <- as.character(sample(c(2, 5, 7, 8), 50, replace = TRUE)) # Example character vector

Our string consists of the four character values 2, 5, 7 & 8:

x                                                            # Print example vector to R console

How to Convert Character to Numeric Example Vector

Graphic 1: Example Character String Printed to the RStudio Console

Now, we can continue with the important part – How to convert this character string to numeric?

No Problem:

x_num <- as.numeric(x)                                 # Convert string to numeric in R
x_num                                                  # Print converted x to the console
# 8 7 5 8 2 5 2 5 2 7 7 7 7...

That’s basically how to apply the as.numeric function in R. However, if you need more explanations, have a look at the following video of my YouTube channel. I’m explaining the topics of this post in the video.

Convert All Characters of a Data Frame to Numeric

As you have seen, to convert a vector or variable with the character class to numeric is no problem. However, sometimes it makes sense to change all character columns of a data frame or matrix to numeric.

Consider the following R data.frame:

x1 <- c("5", "2", "7", "5")                            # Character
x2 <- c("77", "23", "84", "11")                        # Another character
x3 <- as.factor(c("4", "1", "1", "8"))                 # Factor
x4 <- c(3, 3, 9, 7)                                    # Numeric
 
data <- data.frame(x1, x2, x3, x4,                     # Create data frame
                   stringsAsFactors = FALSE)
sapply(data, class)                                    # Print classes of all colums
#          x1          x2          x3          x4 
# "character" "character"    "factor"   "numeric"

Data Frame with Character, Factor & Numeric Columns

Table 1: Example Data Frame with Different Variable Classes

With the following R code, you are able to recode all variables – no matter which variable class – of a data frame to numeric:

data_num <- as.data.frame(apply(data, 2, as.numeric))  # Convert all variable types to numeric
sapply(data_num, class)                                # Print classes of all colums
#        x1        x2        x3        x4 
# "numeric" "numeric" "numeric" "numeric"

However, in many situations it is better to convert only character columns to numeric (i.e. not column X3, since this column should be kept as factor). You could do that with the following code in R:

char_columns <- sapply(data, is.character)             # Identify character columns
data_chars_as_num <- data                              # Replicate data
data_chars_as_num[ , char_columns] <- as.data.frame(   # Recode characters as numeric
  apply(data_chars_as_num[ , char_columns], 2, as.numeric))
sapply(data_chars_as_num, class)                       # Print classes of all colums
#        x1        x2        x3        x4 
# "numeric" "numeric"  "factor" "numeric"

Further Example Videos & Resources: How to Change Variable Types

Are you looking for more explanations on how to convert different data types? Have a look at this video on my YouTube channel:

Further examples needed? Have a look at the following R Programming tutorial of the YouTube Channel LearnR. The speaker discusses different data transformations from one data class to another.

Don
August 5, 2020 1:14 pm

I spent all afternoon scrolling through Stack Overflow trying to figure out how to do this. Once I found it on your site, it took 5 minutes. Thank you!!

Reply
- Joachim
  August 17, 2020 6:54 am
  
  Wow that’s an amazing feedback! Thanks Don 🙂
  
  Reply
Don
August 5, 2020 2:44 pm

How can I keep this from converting my rowname chars to numeric?

Reply
- Joachim
  August 17, 2020 7:01 am
  
  Hey Don,
  
  You may convert only a subset of your data frame to numeric. Have a look here for more info.
  
  Regards,
  
  Joachim
  
  Reply
Larry Cordell
September 2, 2020 5:37 pm

I just had 3 variables in a 17-variable data set to convert from character to numeric. This didn’t help at all I’m afraid.

Reply
- Joachim
  September 3, 2020 6:15 am
  
  Hi Larry,
  
  Could you explain why it didn’t help? Maybe we can figure out a solution for your problem 🙂
  
  Regards,
  
  Joachim
  
  Reply
Vaibhav Datta
September 12, 2020 7:29 pm

Hi. How to stop getting the “Coerced to NA” warning?

Reply
- Joachim
  September 14, 2020 6:02 am
  
  Hi Vaibhav,
  
  This typically happens when your characters are not representing numbers (e.g. “5!”, “1,55”, “seven”). Have you checked how your data is formatted before converting it?
  
  Regards,
  
  Joachim
  
  Reply
  - Alexandre Jordao Chauque
    May 3, 2021 8:23 am
    
    This comment saved my day. I spent almost 3 hours trying to sort out similar problem with my data and then I found it to solve it. Thank you Joachim
    
    Reply
    - Joachim
      May 3, 2021 8:28 am
      
      Hey Alexandre,
      
      That’s great to hear! Glad it helped 🙂
      
      Regards
      
      Joachim
      
      Reply
  - Temitope
    December 10, 2022 2:21 am
    
    Good Day, my issue is i am getting (as.numeric(dailyActivity_merged$ActivityDate)) :
    NAs introduced by coercion, i am a beginner please how do i format my data before converting it. I
    want to convert a date column which is a charter to numeric and also change the format to yyyymmdd .
    
    Reply
    - Cansu (Statistics Globe)
      December 13, 2022 3:30 pm
      
      Hello Temitope,
      
      Are you sure that the class of your Activity Date column is a character? Could you please check it first with class(dailyActivity_merged$ActivityDate)? Besides, what is the reason that you would like to use it in a numeric class, not in a date class?
      
      Regards,
      Cansu.
      
      Reply
Busola
September 15, 2020 6:17 pm

I have factors with levels and labels; how do I convert them to numeric

Reply
- Joachim
  September 16, 2020 5:49 am
  
  Hey Busola,
  
  I have another article on converting factors to numeric: https://statisticsglobe.com/how-to-convert-a-factor-to-numeric-in-r/
  
  Let me know if this solved your problem.
  
  Regards,
  
  Joachim
  
  Reply
SHRINIVAS
November 28, 2020 6:29 am

Hello Jo

I have a data frame which is all in text. Recent Kaggle competition (https://www.kaggle.com/c/kaggle-survey-2020)

While converting from Character to Numeric, I am having problem of NA Coercion

So I saw your Youtube Video and then looked up the above tutorial. Here a2 is the original data set of Kaggle imported as Header F.

However, I am still getting NA coercion. Please help

data <- data.frame(a2$V2, a2$V3, a2$V4, a2$V5, stringsAsFactors = FALSE)
sapply(data, class)

a2.V2 a2.V3 a2.V4 a2.V5
"numeric" "character" "character" "character"

data1 sapply(data_num, class)
Error in lapply(X = X, FUN = FUN, …) : object ‘data_num’ not found

> sapply(data1, class)

a2.V2 a2.V3 a2.V4 a2.V5
“numeric” “numeric” “numeric” “numeric”

> head(data1, 6)

a2.V2 a2.V3 a2.V4 a2.V5
1 NA NA NA NA
2 NA NA NA NA
3 NA NA NA NA
4 NA NA NA NA
5 NA NA NA NA
6 NA NA NA NA
>
> data sapply(data, class)
a2.V2 a2.V3 a2.V4 a2.V5
“numeric” “character” “character” “character”
> data1 sapply(data_num, class)
Error in lapply(X = X, FUN = FUN, …) : object ‘data_num’ not found
> sapply(data1, class)
a2.V2 a2.V3 a2.V4 a2.V5
“numeric” “numeric” “numeric” “numeric”
> head(data1,6)
a2.V2 a2.V3 a2.V4 a2.V5
1 NA NA NA NA
2 NA NA NA NA
3 NA NA NA NA
4 NA NA NA NA
5 NA NA NA NA
6 NA NA NA NA
>

Reply
- Joachim
  November 28, 2020 8:16 am
  
  Hey Shrinivas,
  
  I hope you doing fine?
  
  Regarding your question, I would have a look at two things:
  
  1) The NA coercion problem usually appears, because the character numbers are not formatted properly. More info: https://statisticsglobe.com/warning-message-nas-introduced-by-coercion-in-r
  
  2) It seems like you are trying to use a data frame that does not exist in your workspace. For that reason you get the error “object ‘data_num’ not found”. Please check if this data frame really exists.
  
  I hope that helps!
  
  Joachim
  
  Reply
Leo
January 4, 2021 2:59 pm

Dear Joachim,

thanks for your great videos and website. This time however, I wasn’t able to solve my problem as its more specific but I hope you can bring more light into this:
I have a given column (CV that I want to test in a ANCOVA) in my data set which contains numbers similar to this “-,038040659585351” and its structure is character by default. Now nothing has worked to convert this into numeric. It either got changed into NAs or a whole lot of different numbers.

Do you happen to have an idea to convert it into numeric?

Many thanks in advance!

Leo

Reply
- Joachim
  January 5, 2021 6:47 am
  
  Hey Leo,
  
  Thanks a lot for the nice feedback! I’m happy to hear that you like my content 🙂
  
  Regarding your question, I would try the following:
  
  Step 1) Make sure that the column is a character (not a factor) as explained here: https://statisticsglobe.com/convert-factor-to-character-in-r
  
  Step 2) Use the gsub function to replace all non-numeric symbols and letters in your data. This depends a bit on the exact structure of your data, however, you can find a detailed tutorial here: https://statisticsglobe.com/sub-gsub-r-function-example
  
  Step 3) Convert your column to numeric as shown in this tutorial.
  
  Let me know if it worked 🙂
  
  Joachim
  
  Reply
Elisa
January 16, 2021 10:12 pm

Hi Joachim,

I really appreciate your examples. I am struggling to figure out how to convert a character field with the values Y or N to a Numeric where the Y=1 and the N=0. Similarly, I need to understand how to change Education Levels e.g., Sixth grade to 6, 9th grade to 9th, High School Diploma to 12, … PhD to 22 etc.

Any help you can provide with this is much appreciated.

Thanks again.

Reply
- Joachim
  January 18, 2021 7:23 am
  
  Hey Elisa,
  
  Thank you for the nice feedback!
  
  Regarding your question, please follow these steps:
  
  1) Make sure that your column has the character class: https://statisticsglobe.com/convert-factor-to-character-in-r
  
  2) Replace your values: https://statisticsglobe.com/r-replace-value-of-data-frame-variable-dplyr-package
  
  3) Convert your column to numeric as shown in this tutorial.
  
  Regards,
  
  Joachim
  
  Reply
Thomas Oliver Mérő
March 26, 2021 2:44 pm

Hi Joachim,

I was trying to convert some character variables into numeric variables accrding to your recipe. As you can see I managed to convert them. However, when I checked the entire data set with str() function, its clearly seen that, the variables are still presented as character variables. And they still behave strange when I run them in LM. How can I convert every variable containing numbers into numeric variables in R. Please see below:

> F.BEHAV M.BEHAV OVERALL.SUCCESS i data_num sapply(data_num, class)
M.BEHAV F.BEHAV OVERALL.SUCCESS
“numeric” “numeric” “numeric”

When using the str() function:

> str(GRW_ANALYSIS)
‘data.frame’: 94 obs. of 42 variables:
$ LOCATION : chr “Bager” “Bager” “Kigyos kolut” “Pista” …
$ HABITAT : chr “MP” “MP” “SC1” “MP” …
$ HABITAT.ID : int 1 1 4 1 5 3 5 1 3 5 …
$ YEAR : int 2008 2009 2009 2009 2009 2009 2010 2010 2010 2010 …
$ NO.OF.NESTS : int 17 11 10 5 9 10 8 39 16 14 …
$ NEST.DENSITY…HA : chr “13,0769230769231” “8,46153846153846” “10” “7,142857” …
$ FIRST.FLEDGL.JULIAN : int 45 52 52 44 58 NA 89 49 NA 57 …
$ NEST.HEIGHT : chr “77,1” “123,4” “102,9” “114,848484848485” …
$ WATER.DEPTH : chr “59,9” “7,4” “0” “5,78125” …
$ PROP.MANAG : int 0 85 0 0 0 20 100 0 15 0 …
$ JULIAN.DATE.MANAG : int 0 300 0 0 0 310 290 0 295 0 …
$ PEARCH.AVAIL.10m : int 1 1 2 4 3 4 3 1 4 2 …
$ PRECIP : chr “190,6” “184” “184” “184” …
$ PRECIP.DAY : chr “6,35333333333333” “5,75” “5,75” “5,75” …
$ AVG.MAX.DAILY.PREC : chr “24,6666666666667” “18,9” “18,9” “18,9” …
$ PREC.DAYS…10mm : int 4 5 5 5 5 5 10 10 10 10 …
$ MEAN.TEMP : chr “20,57” “20,3” “20,3” “20,3” …
$ PRECIP.AUG.APR : chr “553,5” “377,9” “377,9” “377,9” …
$ MEAN.TEMP.ANN : chr “12,4” “12,3” “12,3” “12,3” …
$ WIND..6B : int 21 29 29 29 29 29 33 33 33 33 …
$ WIND..8B : int 2 4 4 4 4 4 7 7 7 7 …
$ PROP.PRED.NESTS : chr “0,125” “0” “0,2” “0” …
$ PROP.ABAND.NESTS : chr “0,25” “0,273” “0,2” “0” …
$ PROP.PARA.NESTS : chr “0,059” “0” “0” “0,4” …
$ PROP.SUCC.NESTS : chr “0,588” “0,727” “0,6” “0,6” …
$ MIN.EGGS : int 2 1 3 3 2 1 1 1 2 2 …
$ MAX.EGGS : int 7 5 5 5 5 5 4 6 5 6 …
$ MEAN.EGGS : chr “3,353” “4,09” “4” “4,2” …
$ MIN.FLEDGLING : int 1 2 3 2 2 0 4 1 0 3 …
$ MAX.FLEDGLING : int 5 4 4 5 4 0 4 5 0 4 …
$ MEAN.FLEDGLING : chr “3,1” “3,5” “2” “2,2” …
$ MEAN.EGG.LOSSES : chr “0,176” “0,909” “1” “0,8” …
$ MEAN.NESTLING.LOSSES: chr “0,882” “0,364” “1” “1” …
$ MEAN.UNHATCHED : chr “0,471” “0,273” “0” “0,2” …
$ EGG.DSR : chr “0,9892” “0,9822” “0,9659” “0,9813” …
$ NESTLING.DSR : chr “0,9636” “0,992” “0,9629” “0,97” …
$ Z.VALUE : chr “2,31” “-1,43” “0,14” “0,69” …
$ P.VALUE : chr “0,0211” “0,1528” “0,8911” “0,4851” …
$ HATCHING.RATE : chr “0,851851851851852” “0,914285714285714” “1” “0,941176470588235” …
$ M.BEHAV : chr “” “” “” “” …
$ F.BEHAV : chr “” “” “” “” …
$ OVERALL.SUCCESS : chr “0,479219917494639” “0,669300512211985” “0,418948107828922” “0,520659094344318” …

Thank you, sincerely,
Thomas

Reply
- Joachim
  March 26, 2021 3:23 pm
  Hi Thomas,
  
  It seems like you have stored the numeric variables in a new data.frame called data_num, but then you are running the str() function on your old data.frame GRW_ANALYSIS. You can convert the variables in GRW_ANALYSIS one-by-one using the following R code:
  GRW_ANALYSIS$M.BEHAV <- as.numeric(GRW_ANALYSIS$M.BEHAV) GRW_ANALYSIS$F.BEHAV <- as.numeric(GRW_ANALYSIS$F.BEHAV) GRW_ANALYSIS$OVERALL.SUCCESS <- as.numeric(GRW_ANALYSIS$OVERALL.SUCCESS)
  I hope that helps!
  
  Joachim
  Reply
Suju
May 17, 2021 5:33 pm

Thank you. Your page is really helpful.

Reply
- Joachim
  May 18, 2021 6:52 am
  
  Thanks a lot Suju, that’s really great to hear! 🙂
  
  Reply
SHRINIVAS DHARMADHIKARI
June 13, 2021 2:53 am

Hello Joe
Hope you are doing well and keeping safe.
Your website is my frequent stop for any debugging issue. Compliments on these monumental efforts.

I keep encountering the following problem in R

Whenever in Data Frame there is a column with is actually numeric but R has treated it as String data.

There are two ways I approach and both fail

If I write StringAsFactor= T , the strings column gets converted to factor but then subsequently if I try to convert it to numeric I get all elements in that column coerced to NA

If I write StringAsFactor=F the strings column remains as it is but then subsequently if I try to convert it to numeric I get all elements in that column coerced to NA

Please Guide

SHRINIVAS DHARMA

Reply
- Joachim
  June 14, 2021 6:44 am
  
  Hey Shrinivas,
  
  Nice to hear from you, I’m good and you? 🙂
  
  Thanks a lot for the awesome feedback, it’s really nice to see that you are still reading my website!
  
  Regarding your question:
  
  Usually, it should be possible to convert you string to numeric values using the as.numeric function. However, it seems like your strings are not formatted as proper numbers. Could you share some example values of your data?
  
  Regards
  
  Joachim
  
  Reply
Joshua
October 31, 2021 1:35 am

Good Afternoon Joachim,
I hope your day is going well. I’m having an issue converting a character column from excel data that I read in. My data column involves negative numbers and when I use the following code it does successfully convert to numeric, but the negative numbers are replaced with NA.
chars <- sapply(prob2, is.character)
as.data.frame(apply(prob2[chars],2,as.numeric))
this is the code I used. I'm new to R and could use any help. I also don't know why I had to put a 2 before the as.numeric.
Please Help,
Joshua Fallo

Reply
- Joachim
  November 2, 2021 8:49 am
  
  Hey Joshua,
  
  Thanks, I’m fine, and you? 🙂
  
  It seems like your negative numbers are not formatted correctly. Is there maybe a space in those character strings (i.e. “- 1”)? In this case, you would have to remove this space before converting your data to numeric.
  
  I hope that helps!
  
  Joachim
  
  Reply
Clara
July 4, 2022 9:34 am

hi Joachim,

Thanks for the tutorial. I am having the issue at third step, will you be able to help please?
Here are the details:

> sapply(data2, class) # Print classes of all colums
ID conc value
“factor” “character” “numeric”
> char_columns data_chars_as_num data_chars_as_num[ , char_columns] sapply(data_chars_as_num, class) # Print classes of all colums
ID conc value
“factor” “character” “numeric”

Thanks.

Regards,
Clara

Reply
- Joachim
  July 8, 2022 6:46 pm
  
  Hey Clara,
  
  Thanks for the comment! Could you please explain what your issue exactly is? Is there an unexpected output, or did you get any error messages?
  
  Regards,
  Joachim
  
  Reply
Layefa
November 7, 2022 8:48 pm

Hi Joachim,
I ran this:
compare_df_cols(oct21, nov21, dec21, jan22, feb22, mar22, apr22, may22, jun22,
jul22, sept22, return = “mismatch”)
and got this result:
column_name oct21 nov21 dec21 jan22 feb22 mar22 apr22 may22
1 end_lat numeric numeric numeric numeric numeric character numeric numeric
2 end_lng numeric numeric numeric numeric numeric character numeric numeric

What code do I run to change that of “mar22” to numeric as the others? Dont know if this is the reason why my rbind and bind_rows don’t work.

Reply
- Joachim
  November 14, 2022 12:53 pm
  
  Hi Layefa,
  
  I apologize for the delayed reply. I was on a long holiday, so unfortunately I wasn’t able to get back to you earlier. Do you still need help with your syntax?
  
  Regards,
  Joachim
  
  Reply
John
November 30, 2022 5:34 pm

Dear Joachim,

John here, Hope you are doing good!
I need a help, I have 3000 rows dataset, some of the columns have values like Simple, Medium and High. Would you be able to help to me to convert the values into Simple is 1, Medium is 2 and High is 3. so that I will be able to do ggplot visualization using the data. Thank you so much for your tutorial, It is really helping me a lot.

Best Regards,
John

Reply
- Cansu (Statistics Globe)
  December 14, 2022 10:07 am
  Hello John,
  
  Sorry for the late response, we were a bit busy with some new content creation. I hope you are doing well 🙂 I am afraid that you can not convert a string like “Simple”, etc. to a number directly via the method shown in this tutorial. What you can do is a simple data manipulation as follows.
  data$column[data$column=="Simple"]<-1 data$column[data$column=="Medium"]<-2 data$column[data$column=="High"]<-3 head(data)
  For further content on data manipulation, you can check our tutorial about data manipulation!
  
  Regards,
  Cansu
  Reply
Rafael Alexandrino
December 12, 2022 7:38 pm

Hi! Thanks a lot. I just didn’t manage to do one thing. My data just starts at the 4th row, so I wanted to transform to numeric values skipping the first 3 rows. Can I do that? If not, what is the solution?

Reply
- Cansu (Statistics Globe)
  December 13, 2022 4:24 pm
  
  Hello Rafael,
  
  To the best of my knowledge, a data column can only have one class. Is there any reason that your data starts from the 4th row? Then maybe I can give an alternative solution accordingly.
  
  Regards,
  Cansu
  
  Reply
Lucia
January 10, 2023 3:24 pm

Hello,
I could change my factor to numeric, but all the numbers that have decimals, when I charge the data the program write NA, it only read the numbers that doesn’t have decimals. If someone could tell me what can i do please.

Reply
- Cansu (Statistics Globe)
  January 10, 2023 4:05 pm
  
  Hello Lucia,
  
  It might be something to do with how the decimals are written. It is usually a problem if it is written like 1,2. Instead, it should be like 1.2.
  
  Regards,
  Cansu
  
  Reply
Amy
February 5, 2023 11:17 am

Completely new to R so excuse my lack of understanding

I have got a treatment group which has three levels when imported the column has come up as a character, with the treatments reading “Evit000” “Evit100” and “Evit200”, It will allow conversion to a factor fine and assigns correctly.
However, when looking at conversion to a number- 0,100 and 200 respectively it assigns Na to all columns and will not allow for differentiation between them; is there a way as further down the line I would need to multiply the dose to feed intake which is in a different column.
I noticed this page only allows different assignments when the data is already in a numeric format and has just not been read in correctly.

many thanks

Reply
- Cansu (Statistics Globe)
  February 6, 2023 11:16 am
  Hello Amy,
  
  If I understand you correctly, you want to create a variable with numbers 0, 100, and 200. If so, you should apply a different data manipulation. As you have noticed, you can only convert the data type to numeric as long as your data is in a numeric format but has a factor or character data type.
  data<-data.frame(evit=c("Evit000" , "Evit000", "Evit000" , "Evit100", "Evit100", "Evit200","Evit200","Evit200" )) data # evit # 1 Evit000 # 2 Evit000 # 3 Evit000 # 4 Evit100 # 5 Evit100 # 6 Evit200 # 7 Evit200 # 8 Evit200 data$doses[data$evit=="Evit000"]<-0 data$doses[data$evit=="Evit100"]<-100 data$doses[data$evit=="Evit200"]<-200 data # evit doses # 1 Evit000 0 # 2 Evit000 0 # 3 Evit000 0 # 4 Evit100 100 # 5 Evit100 100 # 6 Evit200 200 # 7 Evit200 200 # 8 Evit200 200
  Regards,
  Cansu
  Reply
Nhi Le
February 20, 2023 2:40 pm

Hi,
I am doing senior projects and i have problem with running variables for multiple linear regression. I want to run heating map (or do correlation) before running ml function. However, I need to convert all 79 variables to numeric. Do you have any advice on this? I think some variables do not make sense to convert to numeric.

Reply
- Cansu (Statistics Globe)
  February 21, 2023 10:21 am
  
  Hello Nhi,
  
  I don’t know your dataset well, but if your variables are not compatible with conversion to numeric, maybe you should consider a different method to evaluate the associations. For instance, the chi-square test. See other alternatives on this page.
  
  Regards,
  Cansu
  
  Reply