用R中的分组和条件根据条件替换一行中的负值

Question

我有一个这样的数据框，

Industry = c("Industry1", "Industry1", "Industry1", "Industry2", "Industry2", "Industry3")
calc = c(-1.2345, 67890, 45678, -9.86544, 32456, 56789)
cy = c(0, 2019, 2017, 2016, 0, 2015)
py = c(0, 2018, 2016, 2015, 0, 2014)

data = data.frame(Industry, calc, cy, py, stringsAsFactors = FALSE)

当 cy = 0 和 py = 0 时，我想通过根据行业分组计算 calc 列的平均值来替换 calc 列中的负值。我尝试了以下方法，

data = data %>% 
  group_by("Industry") %>% 
  mutate(calc = ifelse(cy == 0 & py == 0, summarise(calc = mean(calc)), calc))

但这不起作用。谁能帮我解决这个问题？

Answer 1

你也可以使用replace

Industry = c("Industry1", "Industry1", "Industry1", "Industry2", "Industry2", "Industry3")
calc = c(-1.2345, 67890, 45678, -9.86544, 32456, 56789)
cy = c(0, 2019, 2017, 2016, 0, 2015)
py = c(0, 2018, 2016, 2015, 0, 2014)

data = data.frame(Industry, calc, cy, py, stringsAsFactors = FALSE)
library(tidyverse)

data %>% group_by(Industry) %>%
  mutate(calc = replace(calc, calc <0 & cy == 0 & py == 0, mean(calc[calc >= 0])))
#> # A tibble: 6 x 4
#> # Groups:   Industry [3]
#>   Industry      calc    cy    py
#>   <chr>        <dbl> <dbl> <dbl>
#> 1 Industry1 56784        0     0
#> 2 Industry1 67890     2019  2018
#> 3 Industry1 45678     2017  2016
#> 4 Industry2    -9.87  2016  2015
#> 5 Industry2 32456        0     0
#> 6 Industry3 56789     2015  2014

^{由reprex package (v2.0.0) 于 2021-06-17 创建}

Answer 2

你可以直接变异，不用summarise()。

data %>% 
  group_by(Industry) %>% 
  mutate(
    calc = if_else(calc < 0 & cy == 0 & py == 0, mean(calc[calc > 0]), calc)
  ) %>%
  ungroup()

# # A tibble: 6 x 4
#   Industry      calc    cy    py
#   <chr>        <dbl> <dbl> <dbl>
# 1 Industry1 56784        0     0
# 2 Industry1 67890     2019  2018
# 3 Industry1 45678     2017  2016
# 4 Industry2    -9.87  2016  2015
# 5 Industry2 32456        0     0
# 6 Industry3 56789     2015  2014