Question

I’m having a simple issue that i couldn’t solve with the for loop.
I calculate and save the mean of a column each time inside my for loop.

In the first loop, the mean is saved to the variable. but in the second loop the mean of column2 will replace the mean of column1. This is a very basic question but I have no idea how to do that.

#read.table(file1)
for (x in 1:100){
f <- mean(file1[,x])
}

I want to save the mean values for all columns in 1 variable (assume it is called “f”).

f <- c(meancol1, meancol2, meancol3... meancol100)

Any ideas?

Answer 1

Even simpler you could use colMeans function. Here’s an example

> set.seed(001)  # Generating some random data
> file1 <- data.frame(matrix(rnorm(50, 100, 5), ncol=5))
> file1 # this is how the artificial data.frame should look like
          X1        X2        X3        X4        X5
1   96.86773 107.55891 104.59489 106.79340  99.17738
2  100.91822 101.94922 103.91068  99.48606  98.73319
3   95.82186  96.89380 100.37282 101.93836 103.48482
4  107.97640  88.92650  90.05324  99.73097 102.78332
5  101.64754 105.62465 103.09913  93.11470  96.55622
6   95.89766  99.77533  99.71936  97.92503  96.46252
7  102.43715  99.91905  99.22102  98.02855 101.82291
8  103.69162 104.71918  92.64624  99.70343 103.84266
9  102.87891 104.10611  97.60925 105.50013  99.43827
10  98.47306 102.96951 102.08971 103.81588 104.40554
> colMeans(file1)  # this part computes the means for each column without a 'for' loop
       X1        X2        X3        X4        X5 
100.66101 101.24422  99.33163 100.60365 100.67068 

Take a look at ?colMeans

If you have non-numeric columns you can skip them autotically using sapply inside colMeans for example:

set.seed(001)  # Generating some random data
file1 <- data.frame(matrix(rnorm(50, 100, 5), ncol=5))
# Creating three columns with non-numeric data
factors <- data.frame(matrix(sample(letters, 30, TRUE), ncol=3))

file1 <- cbind(factors, file1) # this is your data.frame
colnames(file1) <- paste0('Col.', 1:ncol(file1)) # set some colnames
file1 # this is the data.frame to work with
> colMeans(file1[sapply(file1, is.numeric)])# colmeans for only those numeric cols
    Col.4     Col.5     Col.6     Col.7     Col.8 
100.65027 101.52467 102.04486  99.14944 100.23847