李家翔 2019-04-03
suppressMessages(source(here::here("R/load.R")))
df <- tibble(A = as.character(c(1,1,1,0,0)), B = c(NA,NA,10,NA,20), D = c(NA,10,10,20,20))
rs <- tibble(col_names = c('B','D'), na_num_1 = c(2,1),
na_num_0 = c(1,0), na_per_1 = c(2/3, 1/2),
na_per_0 = c(1/3,0))
df
## # A tibble: 5 x 3
## A B D
## <chr> <dbl> <dbl>
## 1 1 NA NA
## 2 1 NA 10
## 3 1 10 10
## 4 0 NA 20
## 5 0 20 20
rs
## # A tibble: 2 x 5
## col_names na_num_1 na_num_0 na_per_1 na_per_0
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 B 2 1 0.667 0.333
## 2 D 1 0 0.5 0
df %>%
group_by(A) %>%
mutate_all(is.na) %>%
summarise_all(sum) %>%
# 前面和你差不多，根据 rs 需要转置
column_to_rownames('A') %>%
t() %>%
# 加入比例
{cbind(.,./rowSums(.))} %>%
# 命名
`colnames<-`(paste0(colnames(.),
c(rep("num",2),rep("num",2))))
## `mutate_all()` ignored the following grouping variables:
## Column `A`
## Use `mutate_at(df, vars(-group_cols()), myoperation)` to silence the message.
## 0num 1num 0num 1num
## B 1 2 0.3333333 0.6666667
## D 0 1 0.0000000 1.0000000