替换在R中数据帧任一端找到的连续零

Question

我需要用 NA 替换数据帧的第一列和最后一列中的任何零，但是，当替换第一个/最后一个零时，我还需要替换该特定行中存在的任何连续零。给定示例数据框:

df <- data.frame(a = c(1,0,1,0,1,1,1,0,1,1,1),
                 b = c(1,1,1,0,1,1,1,0,1,1,1),
                 c = c(1,0,1,1,1,0,1,0,1,1,1),
                 d = c(1,1,1,0,1,1,1,1,1,1,1),
                 e = c(1,0,1,0,1,1,1,1,1,1,1),
                 f = c(1,1,1,1,1,1,1,1,1,0,1))
df

我需要它返回:

df.result <- data.frame(a = c(1,NA,1,NA,1,1,1,NA,1,1,1),
                        b = c(1,1,1,NA,1,1,1,NA,1,1,1),
                        c = c(1,0,1,1,1,0,1,NA,1,1,1),
                        d = c(1,1,1,0,1,1,1,1,1,1,1),
                        e = c(1,0,1,0,1,1,1,1,1,1,1),
                        f = c(1,1,1,1,1,1,1,1,1,NA,1))
df.result

提前致谢。

Answer 1

另一种方式，避免apply和对行进行操作:

g<-lapply(df,"==",0)
df[do.call(cbind,Reduce("&",g,accumulate=TRUE)) | do.call(cbind,Reduce("&",g,accumulate=TRUE,right=TRUE))]<-NA
identical(df,df.result)
#[1] TRUE

快速基准测试:

docendo<-function(df) {
  idx <- t(apply(df != 0, 1, function(x) cumsum(x) == 0 | rev(cumsum(rev(x)) == 0)))
  df[idx] <- NA
  df 
}

nicola<-function(df) {
  g<-lapply(df,"==",0)
  df[do.call(cbind,Reduce("&",g,accumulate=TRUE)) | do.call(cbind,Reduce("&",g,accumulate=TRUE,right=TRUE))]<-NA
  df
}

lmo<-function(df) {
   reps.first <- max.col(df, ties.method = "first") - 1
   reps.last <- max.col(df, ties.method = "last")
   fill.last <- length(df)-reps.last
   is.na(df[cbind(rep(seq_len(nrow(df))[reps.first > 0], reps.first[reps.first > 0]),
               sequence(reps.first))]) <- TRUE
   is.na(df[cbind(rep(seq_len(nrow(df))[fill.last > 0], fill.last[fill.last > 0]),
               length(df)-(sequence(fill.last) - 1))]) <- TRUE
   df
}
#create a bigger dataset
df<-df[rep(1:nrow(df),each=10000),]
system.time(res<-docendo(df))
#   user  system elapsed 
#  2.088   0.020   2.145
system.time(res2<-nicola(df))
#   user  system elapsed 
#  0.016   0.000   0.017
identical(res,res2)
#[1] TRUE
system.time(res3<-lmo(df))
#   user  system elapsed 
#  0.222   0.000   0.265
identical(res2,res3)
#[1] TRUE