如何获得具有特定数据的行数?

which 可以在任意数量的列上工作:

# example data
df <- data.frame(
    col1 = c("X", "X", "X", "X", "Z"),
    col2 = c("X", "X", "Z", "X", "X"),
    stringsAsFactors = FALSE
)

# how to get rows without X
unique(sapply(df, function(x) which(x != "X")))

这将起作用:

library(dplyr)
filter(df, (COL1!="X" & COL2!="X"))

这里有一个 tidyverse 可以处理任意数量列的解决方案

library(tidyverse)

df <- tibble(
 col1 = c("X", "X", "X", "X", "Z"),
 col2 = c("X", "X", "Z", "X", "X"),
 col3 = c("X", "X", "Z", "Z", "X"),
)

filter_all(df, any_vars(. != "X"))

返回:

# A tibble: 3 x 3
 col1  col2  col3
  <chr> <chr> <chr>
  1     X     Z     Z
  2     X     X     Z
  3     Z     X     X

下面是一个baser解决方案,它可以用于任意数量的列-

which(rowSums(df == "X") != ncol(df))

你很接近,你想用 sapply 带基元函数 !=

sapply(dat, `!=`, "X")
#      COL1  COL2
#[1,] FALSE FALSE
#[2,] FALSE FALSE
#[3,] FALSE  TRUE
#[4,] FALSE FALSE
#[5,]  TRUE FALSE

如果你需要行号,把它包起来 which ,带参数 arr.ind 设置为 TRUE .

which(sapply(dat, `!=`, "X"), arr.ind = TRUE)
#     row col
#[1,]   5   1
#[2,]   3   2

编辑。

发布了几种解决方案,下面是比较测试。

功能 RuiJaap 是 Rui Jaap评论中的建议。

Rui <- function(DF, value = "X")
  which(sapply(DF, `!=`, value), arr.ind = TRUE)

DanY <- function(DF, value = "X")
  unique(sapply(DF, function(x) which(x != value)))

Jaap <- function(DF, value = "X")
  which(!!rowSums(DF != value))

RuiJaap <- function(DF, value = "X")
  which(dat != "X", arr.ind = TRUE)

library(ggplot2)
library(microbenchmark)

mb1 <- microbenchmark(Rui = Rui(dat),
                      RuiJaap = RuiJaap(dat),
                      Jaap = Jaap(dat),
                      DanY = DanY(dat),
                      times = 1e3)

mb2 <- microbenchmark(Rui = Rui(dat2),
                      RuiJaap = RuiJaap(dat2),
                      Jaap = Jaap(dat2),
                      DanY = DanY(dat2),
                      times = 1e2)

autoplot(mb1)
autoplot(mb2)

对于小型数据集 DanY 鲁雅普 是最快的。

数据。

dat <- read.table(text = "
          COL1         COL2 
1          X            X
2          X            X
3          X            Y
4          X            X
5          Z            X                  
", header = TRUE)


set.seed(1)
dat2 <- matrix("X", nrow = 20, ncol = 100)
dat2[sample(2000, 100)] <- "Y"
dat2 <- as.data.frame(dat2)