如何进行多次除法并将余数存储在 R 中的新变量中？

Question

我有一个数据框，每个数据点的结构如下：ID、度量、时间标记

ID    measure   timemark   
001   12         15    
003   3          13            
004   365        0                   
003   1          13

ID是一个人唯一的study ID，measure是那个人当时使用某项服务的天数，timemark是0到51之间的数字，表示一年有52周x

现在我想创建 52 列的数据框，每一列都包含他们那一周在服务中花费的天数（因此每周的最大天数应该是 7 天）。对于每个人，他们可以在一个时间点有多个条目。从这个意义上讲，总天数应该是两行的总和。

所以我想把它变成这样：

ID    ... week13 week14 week15 week 16   
001   ... 0      0      7      5        
003   ... 4      0      0      0            
004   ... 7      7      7      7

我一直在纠结里面的逻辑，猜想应该是和度量的商余数有关，但是一直推不开。有人可以帮忙吗？

Answer 1

我们可以先为每个 ID 和 timemark 以及 sum 和 measure 值创建一行。我们创建一个列表，将 measure 分成 7 个步长及其余数。使用 unnest_longer 我们获取长格式数据并创建 timemark 列附加周数，最后 spread 宽格式数据。

library(dplyr)
library(tidyr)

df %>%
  group_by(ID, timemark) %>%
  summarise(measure = sum(measure)) %>%
  mutate(measure = list(c(rep(7, floor(measure/7)), measure %% 7))) %>%
  unnest_longer(measure) %>%
  mutate(timemark = paste0('week', first(timemark) + 0:(n() - 1))) %>%
  slice(1:pmin(n(), 52)) %>%
  mutate(timemark = factor(timemark, levels = paste0('week', 0:51))) %>%
  spread(timemark, measure)
  #Or using pivot_wider in new tidyr
  #pivot_wider(names_from = timemark, values_from = measure)


# A tibble: 3 x 53
# Groups:   ID [3]
#     ID week0 week1 week2 week3 week4 week5 week6 week7 week8 week9 week10 week11 week12 week13 week14 week15 week16
#  <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
#1     1    NA    NA    NA    NA    NA    NA    NA    NA    NA    NA     NA     NA     NA     NA     NA      7      5
#2     3    NA    NA    NA    NA    NA    NA    NA    NA    NA    NA     NA     NA     NA      4     NA     NA     NA
#3     4     7     7     7     7     7     7     7     7     7     7      7      7      7      7      7      7      7
# … with 35 more variables: week17 <dbl>, week18 <dbl>, week19 <dbl>, week20 <dbl>, week21 <dbl>, week22 <dbl>,
#   week23 <dbl>, week24 <dbl>, week25 <dbl>, week26 <dbl>, week27 <dbl>, week28 <dbl>, week29 <dbl>, week30 <dbl>,
#   week31 <dbl>, week32 <dbl>, week33 <dbl>, week34 <dbl>, week35 <dbl>, week36 <dbl>, week37 <dbl>, week38 <dbl>,
#   week39 <dbl>, week40 <dbl>, week41 <dbl>, week42 <dbl>, week43 <dbl>, week44 <dbl>, week45 <dbl>, week46 <dbl>,
#   week47 <dbl>, week48 <dbl>, week49 <dbl>, week50 <dbl>, week51 <dbl>

数据

df <- structure(list(ID = c(1L, 3L, 4L, 3L), measure = c(12L, 3L, 365L, 
1L), timemark = c(15L, 13L, 0L, 13L)), class = "data.frame", row.names = c(NA, -4L))

Answer 2

我想留下我为你努力过的东西。首先，我使用 expand() 创建了一个主数据框，其中包含每个 ID 的 ID 和 timemark 的所有组合。然后，我通过以下方式创建了 result。我通过 ID 和 timemark 定义了组并总结了度量。然后，我确定了我需要多少周（行）才能扩展第一个 mutate() 中的结果。然后，我使用 splitstackshape 包中的 expandRows() 扩展了数据框。然后，我更新了 timemark 中的数字，使第二个 mutate() 中的周数正确。然后，我进行了一些计算以分配每周的天数。 lag(measure - 7 * row_number(), default = 7) 创建一个向量，其中包含 measure 还剩多少天。我需要使用逻辑条件替换一些数字。对于每个group，当行数为1时，赋值在measure中。当res大于7时，将7赋值给res。（任何大于 7 的数字都是 7，因为每周（行）最多可能需要 7 天。）否则，保留 res 中的原始值。

library(dplyr)
library(tidyr)
library(splitstackshape)

master <- expand(mydf, timemark = 0:51, ID)

group_by(mydf, ID, timemark) %>% 
summarize(measure = sum(measure)) %>% 
ungroup %>% 
group_by(group = 1:n()) %>% 
mutate(nrow = as.integer(measure / 7) + 1) %>% 
expandRows(count = "nrow") %>%
mutate(timemark = first(timemark):(first(timemark) + n() - 1),
       res = lag(measure - 7 * row_number(), default = 7),
       res = case_when(n() == 1 ~ as.numeric(measure),
                       res > 7 ~ 7,
                       TRUE ~ res)) -> result

最后一步是将 result 加入 master。我删除了不必要的列，使数据框变宽，并更新了列名。

left_join(master, result, by = c("ID", "timemark"))%>% 
select(-c(measure, group)) %>% 
spread(key = timemark, value = res, fill = 0) %>%
rename_at(vars(-ID),
          .funs = list(~paste("week", ., sep = "")))


     ID week0 week1 week2 week3 week4 week5 week6 week7 week8 week9 week10 week11 week12 week13 week14 week15 week16
  <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
1     1     0     0     0     0     0     0     0     0     0     0      0      0      0      0      0      7      5
2     3     0     0     0     0     0     0     0     0     0     0      0      0      0      4      0      0      0
3     4     7     7     7     7     7     7     7     7     7     7      7      7      7      7      7      7      7
# … with 35 more variables: week17 <dbl>, week18 <dbl>, week19 <dbl>, week20 <dbl>, week21 <dbl>, week22 <dbl>,
#   week23 <dbl>, week24 <dbl>, week25 <dbl>, week26 <dbl>, week27 <dbl>, week28 <dbl>, week29 <dbl>, week30 <dbl>,
#   week31 <dbl>, week32 <dbl>, week33 <dbl>, week34 <dbl>, week35 <dbl>, week36 <dbl>, week37 <dbl>, week38 <dbl>,
#   week39 <dbl>, week40 <dbl>, week41 <dbl>, week42 <dbl>, week43 <dbl>, week44 <dbl>, week45 <dbl>, week46 <dbl>,
#   week47 <dbl>, week48 <dbl>, week49 <dbl>, week50 <dbl>, week51 <dbl>

数据

mydf <- structure(list(ID = c(1L, 3L, 4L, 3L), measure = c(12L, 3L, 365L, 
1L), timemark = c(15L, 13L, 0L, 13L)), class = "data.frame", row.names = c(NA, 
-4L))

如何进行多次除法并将余数存储在 R 中的新变量中？

How to do several division and store the remainders in new variables in R?

aggregate

r

division