@Hasan, I also recommend stackoverflow.com for all your R questions, it is by far the best and most up to date forum.

Hi Hasan,

You can remove the missing values in the whole data set using "na.omit()".

You should, however, be careful, because it deletes the rows where the NA occurs.

Hope it helps,

Augusto

@AhmedElwan , thank you for your response and interest. When I did your suggestion, it still gives a error like "There were 18 warnings (use warnings() to see them)", then when I checked the errors it says that; In eval(substitute(expr), envir, enclos) : NAs introduced by coercion.

Hi Augusto,

Thank you for your reply. I will try it.

Best,

Oh, I guess it works now! Thank you very much Ahmed! And one more thing, I have to add variance and mean of each rows in a new column and let's say sort the first 500.

Do you think this will works?

df (df

Suggest you read "R for Data Science" to get a basic understanding of modern R before you try to do your analysis. This is an excellent book for beginners and will give you a lot of tools for your work. It's available free here:

http://r4ds.had.co.nz/

Or you can use read_csv2 (which assumes sep=";" and dec=",") and then similar to Ahmed's answer. I like to calculate variance as var = ssq/n - mean^2:

library(tidyverse)

df %

select(-geneID) %>%

mutate(rowcount = rowSums(!is.na(.)),

rowsum = rowSums(. , na.rm=TRUE),

rowmean = rowsum/rowcount,

rowssq = rowSums(.^2 , na.rm=TRUE),

rowvar = rowssq/rowcount - rowmean^2)

fread() is a little faster. Actually using Ahmed's fread() with my mutate() is even faster!

Thank you @Simon for your detailed explanation and recommendation. I am new in R, so I hope the book that you suggested will help me a lot.

I believe that as you full table was a matrix instead of a data frame it will be considered as numeric and should solve your problem

Yes, you are right @Geremy. I fixed it by the way. Thank you for your response.

@Hasan, I also recommend stackoverflow.com for all your R questions, it is by far the best and most up to date forum.

If you are a beginner in R, perhaps it would be easier to try either data.table approach or dplyr (tidyverse) approach. The mixture of their non-base-R syntaxes can be really hard for the beginners. Both are great for data manipulation. Actually, there are long threads for 'data.table vs tidyverse' if you are interested in the details, just check stackoverflow.