将行转换为列并汇总答案

【问题标题】：Transform row to column and Summarise将行转换为列并汇总
【发布时间】：2021-08-05 15:59:09
【问题描述】：

谁能帮我解决这个问题

我有一个数据集，见下文

Merchant	value	Status
paul	5	failed
royal	3	Failed
servy	56	Success
servy	33	success
paul	4	declined
royal	5	declined
royal	6	declined
royal	22	success
paul	11	Requery declined

我要做的是总结状态栏就像使用 excel 数据透视表一样，您可以将行转换为列并进行汇总

我想在 R 中做同样的事情......我很确定必须有一种方法可以使用 dplyr 或 tidyverse 包或任何你知道可以在 R 中完成的包来完成它

总之，我希望我的结果如下表所示。

Merchant	declined	failed	Requerydeclined	Success
paul	4	5	11
royal	11	3		22
servy				89
Total	15	8	11	111

另外我是 R 的初学者

如果我能回答这个请求，我将不胜感激。

谢谢大家

【问题讨论】：

请在此处粘贴dput() 您的数据
同意，不过为了方便将来参考它从剪贴板读取表格df <- read.table("clipboard", sep = '\t', header = T)

标签： r dplyr data-science tidyverse janitor

【解决方案1】：

使用dcast

library(data.table)
dcast(setDT(data), Merchant ~ Status)

或xtabs

xtabs(value ~ Merchant + Status, data)

【讨论】：

【解决方案2】：

你的数据是：

dat <- data.frame(merchant = c("paul", "royal", "servy", "servy", "paul", "royal", "royal", "royal", "paul"),
          value = c(5, 3, 56, 33, 4, 5, 6, 22, 11),
          Status = c("failed", "failed", "success", "success", "declined", "declined", "declined", "success", "Rquery declined"))

下表可以创建：

table <- 
  dat %>% 
  group_by(merchant, Status) %>% 
  summarise(value = sum(value), .groups = "drop") %>% 
  tidyr::pivot_wider(names_from = Status, values_from = value)

你得到总和：

table %>% 
  select(!merchant) %>% 
  summarise(across(everything(), ~ sum(.x, na.rm = T)))

【讨论】：

【解决方案3】：

为此，您需要pivot_wider，

library(tidyverse)

data %>%
   pivot_wider(names_from = Status, values_from = value)

【讨论】：

【解决方案4】：

另一种选择是使用reshape2 包中的dcast() 函数：

library(reshape2)

Merchant<-c("paul", "royal", "servy", "servy", "paul", "royal", "royal", "royal", "paul")
Value<-c(5,3,56,33,4,5,6,22,11)
Status<-c("failed", "failed", "Success", "Success", "declined", "declined", "declined", "Success", "Requery declined")

DF<-data.frame(Merchant=Merchant, Value=Value, Status=Status)

dcast(DF, Merchant~Status, fun.aggregate=sum, value.var="Value")

【讨论】：

【解决方案5】：

tidyverse 确实可以提供帮助，即：

library(dplyr)
library(tidyr) #for spread
library(stringr) #for str_to_title
df %>%
  mutate(Status = str_to_title(Status)) %>% # make failed and Failed the same
  group_by(Merchant, Status) %>%
  summarise(value = sum(value)) %>% #summarise so each merchant and status only has one value value
  spread(key = Status, value = value, fill = 0) #the pivot section

编辑 - spread 现在已经过时，所以这里是 pivot_wider 版本。

df %>%
  mutate(Status = str_to_title(Status)) %>% # make failed and Failed the same
  group_by(Merchant, Status) %>%
  summarise(value = sum(value)) %>% #summarise so each merchant and status only has one value value
  pivot_wider(names_from = Status, values_from = value) #the pivot section

【讨论】：

spread 被pivot_wider 取代！ :-)