英文:
Mutate a new column according to the values of each row
问题
我有以下的玩具数据框。
toy.df <- data.frame(Name = c("group1", "group2", "group3", "group4", "group5", "group6", "group7"),col1 = c("pos", "neg", "NA", "pos","neg", "NA", "pos"),col2 = c("pos", "pos", "NA", "pos","neg","NA", "neg"),col3 = c("pos", "NA", "pos", "NA", "neg", "neg", "neg"))
我想要创建一个新列，检查每一行的所有列的值。如果它们都是"pos"或"NA"，则变为"pos"，如果它们都是"neg"或"NA"，则变为"neg"，如果它们是"pos"、"neg"或"NA"中的任何一个，则变为"both"。
新列看起来如下：
col4 <- c("pos", "both", "pos", "pos","neg", "neg","both")
这是最终的数据框：
Name col1 col2 col3 col4group1 pos pos pos posgroup2 neg pos NA bothgroup3 NA NA pos posgroup4 pos pos NA posgroup5 neg neg neg neggroup6 NA NA neg neggroup7 pos neg neg both
英文:
I have the following toy data frame.
toy.df <- data.frame(Name = c("group1", "group2", "group3", "group4", "group5", "group6", "group7"),col1 = c("pos", "neg", "NA", "pos","neg", "NA", "pos"),col2 = c("pos", "pos", "NA", "pos","neg","NA", "neg"),col3 = c("pos", "NA", "pos", "NA", "neg", "neg", "neg"))
I would like to mutate a new column that check the values of all columns per row. If they are all "pos" or "NA" mutate "pos", if they are all "neg" or "NA" mutate "neg" and if they are "pos" or "neg" or "NA" mutate "both".
The new column looks as follows:
col4 <- c("pos", "both", "pos", "pos","neg", "neg","both")
Here is the final data frame:
Name col1 col2 col3 col4group1 pos pos pos posgroup2 neg pos NA bothgroup3 NA NA pos posgroup4 pos pos NA posgroup5 neg neg neg neggroup6 NA NA neg neggroup7 pos neg neg both
答案1
得分: 3
以下是您要的翻译部分：
"NA"在您的数据框中是字面值"NA"，我们需要使用na_if将其转换为真正的缺失值NA，然后使用case_when为新列分配条件。我们需要在每一行中使用rowwise才能使其在每一行中起作用。case_when中的最后一个TRUE ~ "unknown"捕捉了col1到col3中除了"pos"和"neg"之外的字符串。
我添加了两个条目来展示当所有行都是NA或列中有拼写错误时的行为。
library(dplyr)toy.df %>%rowwise() %>%mutate(across(everything(), ~na_if(.x, "NA")),col4 = case_when(all(is.na(c_across(col1:col3))) ~ NA,all(c_across(col1:col3) == "pos", na.rm = T) ~ "pos",all(c_across(col1:col3) == "neg", na.rm = T) ~ "neg",all(c_across(col1:col3) %in% c("pos", "neg", NA)) ~ "both",TRUE ~ "unknown")) %>%ungroup()# A tibble: 9 × 5Name col1 col2 col3 col41 group1 pos pos pos pos2 group2 neg pos NA both3 group3 NA NA pos pos4 group4 pos pos NA pos5 group5 neg neg neg neg6 group6 NA NA neg neg7 group7 pos neg neg both8 group8 NA NA NA NA9 group9 pos pos typo unknown
数据
toy.df <- structure(list(Name = c("group1", "group2", "group3", "group4","group5", "group6", "group7", "group8", "group9"), col1 = c("pos","neg", "NA", "pos", "neg", "NA", "pos", NA, "pos"), col2 = c("pos","pos", "NA", "pos", "neg", "NA", "neg", NA, "pos"), col3 = c("pos","NA", "pos", "NA", "neg", "neg", "neg", NA, "typo")), class = "data.frame", row.names = c(NA,-9L))
英文:
Since the "NA" in your data frame is literal "NA", we need to turn it into real missing value NA by na_if. Then use case_when to supply the conditions for new column assignment. We need rowwise for it to work in every row. The final TRUE ~ "unknown" in case_when captures strings other than "pos" and "neg" in col1 to col3.
I added two entries to show the behaviour when all rows are NA, or when there's a typo in the columns.
library(dplyr)toy.df %>%rowwise() %>%mutate(across(everything(), ~na_if(.x, "NA")),col4 = case_when(all(is.na(c_across(col1:col3))) ~ NA,all(c_across(col1:col3) == "pos", na.rm = T) ~ "pos",all(c_across(col1:col3) == "neg", na.rm = T) ~ "neg",all(c_across(col1:col3) %in% c("pos", "neg", NA)) ~ "both",TRUE ~ "unknown")) %>%ungroup()# A tibble: 9 × 5Name col1 col2 col3 col4<chr> <chr> <chr> <chr> <chr>1 group1 pos pos pos pos2 group2 neg pos NA both3 group3 NA NA pos pos4 group4 pos pos NA pos5 group5 neg neg neg neg6 group6 NA NA neg neg7 group7 pos neg neg both8 group8 NA NA NA NA9 group9 pos pos typo unknown
Data
toy.df <- structure(list(Name = c("group1", "group2", "group3", "group4","group5", "group6", "group7", "group8", "group9"), col1 = c("pos","neg", "NA", "pos", "neg", "NA", "pos", NA, "pos"), col2 = c("pos","pos", "NA", "pos", "neg", "NA", "neg", NA, "pos"), col3 = c("pos","NA", "pos", "NA", "neg", "neg", "neg", NA, "typo")), class = "data.frame", row.names = c(NA,-9L))
答案2
得分: 1
以下是您要翻译的代码部分：
toy.df$group6 <- apply(toy.df, 1, \(x) {val <- sort(unique(x[2:4]))if (val[1] == "NA") val = val[2:length(val)]if (length(val) == 2) {"both"} else if (val=="pos")"pos"else"neg"})toy.df
out:
Name col1 col2 col3 group61 group1 pos pos pos pos2 group2 neg pos NA both3 group3 NA NA pos pos4 group4 pos pos NA pos5 group5 neg neg neg neg6 group6 NA NA neg neg7 group7 pos neg neg both
英文:
Another way:
toy.df$group6 <- apply(toy.df, 1, \(x) {val <- sort(unique(x[2:4]))if (val[1] == "NA") val = val[2:length(val)]if (length(val) == 2) {"both"} else if (val=="pos")"pos"else"neg"})toy.df
out:
Name col1 col2 col3 group61 group1 pos pos pos pos2 group2 neg pos NA both3 group3 NA NA pos pos4 group4 pos pos NA pos5 group5 neg neg neg neg6 group6 NA NA neg neg7 group7 pos neg neg both
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论