Question

我有以下格式的数据，我正在尝试创建一个新变量，其中包括每个观察的总死亡人数，其中＆＃34;存在＆＃34;表示正在进行的活动：

    Birth1  Death1  Birth2  Death2  Birth3  Death3  Birth4  Death4  Birth5  Death5  Birth6  Death6
  1 1990    present                                     
  2 1984    1986    1986    present                             
  3 1985    1988    1988    present                             
  4 1987    1991    1991    1994    1996    present                     
  5 1987    1989    1989    present

我试过了data$num.deaths <- ifelse(data$death1=="present", 0, 1)，但显然没有得到有多个死亡事件的观察结果。我也试过嵌套ifelse但得到了相同的结果。任何人都能指出我这样快速有效的方法吗？

Answer 1

提取表示给出Deaths的死亡的列，然后在每一行中累加非NA的元素数，而不是空字符串且不等于"present"。没有包使用。

Deaths <- data[grep("Death", names(data))]
rowSums(!is.na(Deaths) & Deaths != "" & Deaths != "present")
## A B C D 
## 0 1 1 2

给出相同结果的最后一行代码的替代方法是检查每行的每个单元格是否有数字，并将每行中的成功相加 - apply将行转换为列，这样我们就可以了使用colSums来做到这一点。

colSums(apply(Deaths, 1, grepl, pattern = "\\d"))

注意：最好在您的问题中显示dput(data)的输出，以明确且可重复地传达您的输入。如果没有它，你所拥有的和答案之间可能存在微小的差异，因此为了重现性，我们将其用作输入（对应于修改前问题中显示的原始输入数据和样本输出）：

data <- 
structure(list(Birth1 = c(1990L, 1984L, 1985L, 1987L), Death1 = c("present", 
"1986", "1988", "1991"), Birth2 = c(NA, 1986L, 1988L, 1991L), 
    Death2 = c("", "present", "present", "1994"), Birth3 = c(NA, 
    NA, NA, 1996L), Death3 = c("", "", "", "present")), .Names = c("Birth1", 
"Death1", "Birth2", "Death2", "Birth3", "Death3"), 
class = "data.frame", row.names = c("A", "B", "C", "D"))

看起来像这样：

> data
  Birth1  Death1 Birth2  Death2 Birth3  Death3
A   1990 present     NA             NA        
B   1984    1986   1986 present     NA        
C   1985    1988   1988 present     NA        
D   1987    1991   1991    1994   1996 present

Answer 2

Here is another option with Reduce and +. We loop through the columns with lapply, convert the elements to binary (0/1) by checking if the elements contain only number of not, and then with Reduce we sum up the corresponding elements of each row.

Reduce(`+`,lapply(data[grep('Death', names(data))], 
                grepl, pattern='^\\d+$'))
#[1] 0 1 1 2

NOTE: The example is taken from the dput output in @G. Grothendieck's post.

R中嵌套的`ifelse`语句

2 个答案: