r - How can I remove outlines from location data -

i have location data of marine species. however, data fall on land , remove them without finding rows belong using ( ps2<-ps1[-c(1,2,3),]) because have large data set. have search clue can't find good. thank you.

here codes

        require(sp)         library(maptools)         require(raster)         data(wrld_simpl)           ps<-read.csv('test.csv')         ps    #-----------------                   lat        lon 1  -32.98000 154.000000 2   36.94625   8.212916 3  -37.81430 -57.479584 4  -19.77236 -40.019028 5  -25.70459 -48.473195 6  -22.47125 -41.859027 7  -28.08153 -48.627082 8   10.56000  39.090000 9   50.50000  50.600000 10  52.50000   5.700000          ps1 <- subset(ps, !is.na(lon) & !is.na(lat))          plot(wrld_simpl, bg='azure2', col='khaki', border='#aaaaaa')         #restore box around map         box()         #plot points         points(ps1$lon, ps1$lat, col='orange', pch=20, cex=0.75)         # plot points again add border, better visibility         points(ps1$lon, ps1$lat, col='red', cex=0.75)     last 3 data fall on land. 

this shows how can use %over% (you can use over function

library(sp) library(raster) library(maptools) data(wrld_simpl)  ps <- matrix( c(-32.98000, 154.000000, 36.94625, 8.212916,-37.81430, -57.479584,-19.77236, -40.019028,-25.70459, -48.473195,-22.47125, -41.859027,-28.08153, -48.627082, 10.56000,  39.090000,50.50000,  50.600000, 52.50000,   5.700000), ncol=2, byrow=true) ps <- ps[,2:1]  sp <- spatialpoints(ps) crs(sp) = crs(wrld_simpl) <- sp %over% wrld_simpl marine <- is.na(i$iso3)  spm <- sp[marine, ]  plot(wrld_simpl) points(ps, pch=20, col='red') points(spm, pch=20, col='blue')