R 按多列统计频次

发布于03月03日

我有多个专栏，里面有风投的资金.

df <- data.frame(Company = c("X", "Y","Z"),
                 Valuation = c("10","20","30"),
                 VC.1 = c("Zeev Ventures","Bedrock Capital","Activant Capital"),
                 VC.2 = c("Bedrock Capital","Activant Capital","Zeev Ventures"))

Company	Valuatiion	VC.1	VC.2
X	10	Zeev Ventures	Bedrock Capital
Y	20	Bedrock Capital	Activant Capital
Z	30	Activant Capital	Zeev Ventures

我需要Calcucalte for each 风投基金提供不同的统计数据.例如

VC	Total Valuatiion	Total Investments
Zeev Ventures	40	2
Bedrock Capital	40	2
Activant Capital	50	2

推荐答案

您可以透视VC.1列和VC.2列，并使用常见的dplyr个函数获得结果:

library(tidyr)
library(dplyr)
pivot_longer(df, cols = c("VC.1", "VC.2"), values_to = "VC") %>%
  group_by(VC) %>%
  summarise(TV = sum(as.numeric(Valuation)), TI = n())

# A tibble: 3 × 3
  VC                  TV    TI
  <chr>            <dbl> <int>
1 Activant Capital    50     2
2 Bedrock Capital     30     2
3 Zeev Ventures       40     2

我建议单独判断pivot_longer()函数调用的结果，以便更好地理解为什么这个旋转步骤解决了在多个列中具有标识符的问题.