我想在大数据帧中生成一列，其中包含其他列的信息。我举了一个非常小的可复制的例子：
tax <- data.frame(
Family = c("Brassicacae", "Pinaceae", "Rosaceae", "Liliaceae"),
Genus = c("NA" ,"Pinus", "NA", "Lilia"),
Species = c("NA" ,"Pinus_sylvestris", "NA", "Calochortus nuttallii"))
我想创建一个名为tax_rank的列，在该列中，您的分类达到的物种将具有值species，但如果您达到的等级比属更高，则值将为genus或family，如以下输出所示：
tax <- data.frame(
Family = c("Brassicacae", "Pinaceae", "Rosaceae", "Liliaceae"),
Genus = c("NA" ,"Pinus", "NA", "Lilia"),
Species = c("NA" ,"Pinus_sylvestris", "NA", "Calochortus nuttallii"),
tax_rank = c("family" ,"species", "family", "species"))
但我想用一个大数据集自动完成，用dplyr可能吗？谢谢
在base R中，可以对非NA值使用max.col，并选择ties.method = "last"以保留最新的非NA值。
names(tax)[max.col(!is.na(tax), ties.method = "last")]
这可以将其转换为dplyr:
library(dplyr)
tax %>%
mutate(tax_rank = names(tax)[max.col(!is.na(tax), ties.method = "last")])
# Family Genus Species tax_rank
# 1 Brassicacae <NA> <NA> Family
# 2 Pinaceae Pinus Pinus_sylvestris Species
# 3 Rosaceae <NA> <NA> Family
# 4 Liliaceae Lilia Calochortus nuttallii Species
数据(注意，我将"NA"转换为NA(
tax <- data.frame(
Family = c("Brassicacae", "Pinaceae", "Rosaceae", "Liliaceae"),
Genus = c(NA ,"Pinus", NA, "Lilia"),
Species = c(NA ,"Pinus_sylvestris", NA, "Calochortus nuttallii"))
首先，数据帧应该包含NA对象，而不是字符：
tax <- data.frame(
Family = c("Brassicacae", "Pinaceae", "Rosaceae", "Liliaceae"),
Genus = c(NA ,"Pinus",NA, "Lilia"),
Species = c(NA,"Pinus_sylvestris", NA, "Calochortus nuttallii"))
那么你想要的列就是下一个
tax %>% mutate(tax_rank = ifelse(!is.na(Species), "species", ifelse(!is.na(Genus), "genus", "family")))
这是输出
Family Genus Species tax_rank
1 Brassicacae <NA> <NA> family
2 Pinaceae Pinus Pinus_sylvestris species
3 Rosaceae <NA> <NA> family
4 Liliaceae Lilia Calochortus nuttallii species
使用base R
tax$tax_rank <- apply(tax, 1, (x) tail(names(x)[!is.na(x)], 1))
-输出
> tax
Family Genus Species tax_rank
1 Brassicacae <NA> <NA> Family
2 Pinaceae Pinus Pinus_sylvestris Species
3 Rosaceae <NA> <NA> Family
4 Liliaceae Lilia Calochortus nuttallii Species