CODEWITHSUNDEEP
rdataframemean
user2420472
user2420472

Reputation: 1177

calculate the mean for each column of a matrix in R

I am working on R in R studio. I need to calculate the mean for each column of a data frame.

 cluster1  // 5 by 4 data frame
 mean(cluster1) //

I got : 

  Warning message:
  In mean.default(cluster1) :
  argument is not numeric or logical: returning NA

But I can use 

  mean(cluster1[[1]])

to get the mean of the first column.

How to get means for all columns ?

Any help would be appreciated.

Upvotes: 73

Views: 307147

Answers (10)

Quinten
Quinten

Reputation: 41603

Another option using the function fmean from the collapse package. Here is a reproducible example:

set.seed(1)
m <- data.frame(matrix(sample(100, 20, replace = TRUE), ncol = 4))
library(collapse)
fmean(m)

Output:

  X1   X2   X3   X4 
47.0 64.4 44.8 67.8

Upvotes: 1

SysEng
SysEng

Reputation: 53

colMeans(A, na.rm = FALSE, dims = 1)

https://stat.ethz.ch/R-manual/R-devel/library/base/html/colSums.html

This is in the base class, so no library is required.

The first answer looks like it is using colMeans from the analytics library which is not available in the R version 4.0.2.

Upvotes: 1

Seyma Kalay
Seyma Kalay

Reputation: 2863

class(mtcars)
my.mean <- unlist(lapply(mtcars, mean)); my.mean



   mpg        cyl       disp         hp       drat         wt       qsec         vs 
 20.090625   6.187500 230.721875 146.687500   3.596563   3.217250  17.848750   0.437500 
        am       gear       carb 
  0.406250   3.687500   2.812500

Upvotes: 2

I Ju Cheng
I Ju Cheng

Reputation: 23

try it ! also can calculate NA's data!

df <- data.frame(a1=1:10, a2=11:20)

df %>% summarise_each(funs( mean( .,na.rm = TRUE)))


# a1   a2
# 5.5 15.5

Upvotes: 2

weijia
weijia

Reputation: 21

You can try this:

mean(as.matrix(cluster1))

Upvotes: 2

Agaz Wani
Agaz Wani

Reputation: 5694

For diversity: Another way is to converts a vector function to one that works with data frames by using plyr::colwise()

set.seed(1)
m <- data.frame(matrix(sample(100, 20, replace = TRUE), ncol = 4))

plyr::colwise(mean)(m)


#   X1   X2   X3   X4
# 1 47 64.4 44.8 67.8

Upvotes: 0

Gonzalo
Gonzalo

Reputation: 191

In case you have NA's:

sapply(data, mean, na.rm = T)      # Returns a vector (with names)   
lapply(data, mean, na.rm = T)      # Returns a list

Remember that "mean" needs numeric data. If you have mixed class data, then use: 

numdata<-data[sapply(data, is.numeric)]  
sapply(numdata, mean, na.rm = T)  # Returns a vector
lapply(numdata, mean, na.rm = T)  # Returns a list

Upvotes: 13

user6376316
user6376316

Reputation:

Another way is to use purrr package 

# example data like what is said above

@A Handcart And Mohair

set.seed(1)
m <- data.frame(matrix(sample(100, 20, replace = TRUE), ncol = 4))


library(purrr)
means <- map_dbl(m, mean)

> means
#  X1   X2   X3   X4 
#47.0 64.4 44.8 67.8

Upvotes: 2

bob
bob

Reputation: 311

You can use 'apply' to run a function or the rows or columns of a matrix or numerical data frame:

cluster1 <- data.frame(a=1:5, b=11:15, c=21:25, d=31:35)

apply(cluster1,2,mean)  # applies function 'mean' to 2nd dimension (columns)

apply(cluster1,1,mean)  # applies function to 1st dimension (rows)

sapply(cluster1, mean)  # also takes mean of columns, treating data frame like list of vectors

Upvotes: 31

A5C1D2H2I1M1N2O1R2T1
A5C1D2H2I1M1N2O1R2T1

Reputation: 193687

You can use colMeans:

### Sample data
set.seed(1)
m <- data.frame(matrix(sample(100, 20, replace = TRUE), ncol = 4))

### Your error
mean(m)
# [1] NA
# Warning message:
# In mean.default(m) : argument is not numeric or logical: returning NA

### The result using `colMeans`
colMeans(m)
#   X1   X2   X3   X4 
# 47.0 64.4 44.8 67.8

Upvotes: 91

Related Questions