Dear All, I think there is a bug in the convolve function of stats package. Running the following: a <- convolve(c(1,1,1,1), 1, type="filter") a the answer is: [1] 1 1 whereas it should be: [1] 1 1 1 1 Looking at the code of convolve, the bug is on line 22 at: [-c(1L:n1, (n - n1 + 1L):n)]/n which is not correct when the second input argument has only one element (n1=0). When I run R.version I get the following output: platform i686-pc-linux-gnu arch i686 os linux-gnu system i686, linux-gnu status major 3 minor 4.0 year 2017 month 04 day 21 svn rev 72570 language R version.string R version 3.4.0 (2017-04-21) nickname You Stupid Darkness Best, George Anastasiou [[alternative HTML version deleted]]