Remove columns with standard deviation of zero

Question

I want to remove all columns with a standard deviation of zero from a data.frame.

This does not work:

  df <- df[, ! apply(df , 2 , function(x) sd(x)==0 ) ]

I get error:

undefined columns selected

UPDATE

I selected Filter as my preferred answer as it also seems to handle NAs, which is very useful.

For example, in

df <- data.frame(v1=c(0,0,NA,0,0), v2=1:5)

the column 'v1' is removed with Filter while the apply methods produce errors.

Thanks to all the other solutions, I learned a lot from them.

UPDATE2:

Those errors given by apply can be fixed by adding na.rm = TRUE to the call to sd like so:

df[, ! apply(df , 2 , function(x) sd(x, na.rm = TRUE)==0 ) ]

Answer 1

use filter:

Filter(function(x) sd(x) != 0, df)

Answer 2

In addition to @grrgrrbla 's and @akrun 's answer using Filter, here is the correct way to do what you originally had in mind:

df <- df[, !sapply(df, function(x) { sd(x) == 0} )]

Or

df <- df[, sapply(df, function(x) { sd(x) != 0} )]

I used sapply() to get a vector which is TRUE when a data frame column have a standard deviation of 0 and FALSE otherwise. Then I subset the original data frame using this vector.