How to calculate mean value for each column ignoring NA

Question

table, data.frame.

Which has 12 columns with variable names and 24 rows df

Like:

Var1 Var2 Var3 Var4 Var12
1     NA   2     3    4
5     6    2     3    3
NA    7    8     NA   4

And I want to calculate the mean for each column while ignoring the Na's For example:

colMeans(df)

And get the result like:

Var1  Var2  Var3  Var4  Var12
 3     6,5   4     3     3,66

I don't want the NA's to be considered in the calculation of Mean.

I tried some methods like na.omit, !is.na, but I don't get the desired result like what I described above.

The calculation of Means is correct but how can I get it like horizontal while getting the col names above each mean number? — Alex Rika, Mar 28 '19 at 10:33
Not sure what you mean by that. Can you update the post with the expected output? — Ronak Shah, Mar 28 '19 at 10:39
Yeah In the expected output you can see what I want. Is each variable name from the column and below the mean value — Alex Rika, Mar 28 '19 at 10:40
It gives me the same expected output as shown when I do `colMeans(df, na.rm = TRUE)`, I have a variable name and it's mean value below it, not sure what is missing. — Ronak Shah, Mar 28 '19 at 10:44
Yeah It gives A single variable name called "means" But I want the variable names of my Data table which are 12 and below each of the 12 variables to be the single mean value. Like Var1-> 3.32 Var2-> 4.55 etc. — Alex Rika, Mar 28 '19 at 10:46

score 1 · Accepted Answer · answered Mar 28 '19 at 10:53

1

For a data.table dt, that looks like this:

dt
   Var1 Var2 Var3 Var4 Var12
1:    1   NA    2    3     4
2:    5    6    2    3     3
3:   NA    7    8   NA     4

You can simply use lapply():

dt[, lapply(.SD, mean, na.rm = TRUE)]

The result is:

   Var1 Var2 Var3 Var4    Var12
1:    3  6.5    4    3 3.666667

answered Mar 28 '19 at 10:53

clemens

How would you syntax the exact thing but for row means? – Alex Rika Mar 28 '19 at 10:57
You could use `dt[, apply(.SD, 1, mean, na.rm = TRUE)]` – clemens Mar 28 '19 at 11:02
for some reason now the dt[, lapply(.SD, mean, na.rm = TRUE)] gives me an error:Error in .subset(x, j) : invalid subscript type 'list' But before I didnnt get an error? – Alex Rika Mar 28 '19 at 11:05
oh ok I fixed it should have converted back to data table – Alex Rika Mar 28 '19 at 11:12

1 Answers1