堆栈内存溢出
  简体   繁体   English

integer 的条件增量基于 R dplyr group_by 中的条件

[英]Conditional increment of an integer based on a condition inside a R dplyr group_by

 原文  2022-04-04 21:30:14  6   1    r/ dplyr

问题描述

I have got a dataset, where a campaign workflow sends emails on daily basis to contacts that meet certain criteria.我有一个数据集，其中活动工作流程每天向满足特定条件的联系人发送电子邮件。

The workflow sends 3 different email communications A, B, and C separated by a number of days and based on recipients interaction events.工作流发送 3 个不同的 email 通信 A、B 和 C，间隔天数并基于收件人交互事件。 A and B email communications have two versions A1,A2 and B1,B2. A和B email通信有两个版本A1,A2和B1,B2。

Recipients can qualify to enter the campaign workflow again in the future.收件人将来有资格再次进入活动工作流程。 The data set has no meta data to indicate a start of a new campaign for each recipient_id .该数据集没有元数据来指示每个recipient_id的新活动开始。 Therefore to analyse the data further, I need to develop two new columns for identifying new campaign within a recipient_id group and based on some logic described below:因此，为了进一步分析数据，我需要开发两个新列来识别recipient_id组中的新活动，并基于下面描述的一些逻辑：

Approach 1 : if step > lead(step) marks an end to a campaign.方法 1 ：如果step > lead(step)标志着活动结束。 OR if step < lag(step) marks the start of a new campaign, therefore increment campaign count by 1.或者，如果step < lag(step)标志着新活动的开始，则将活动计数增加 1。

Approach 2 : if step > lead(step) & lead(date)-date > 14 days marks end of a campaign.方法 2 ：如果step > lead(step) & lead(date)-date > 14天标志着活动结束。 OR if step < lag(step) & date - lag(date) > 14 days marks the start of a new campaign, therefore increment campaign count by 1.或者，如果step < lag(step) & date - lag(date) > 14天标志着新活动的开始，因此将活动计数增加 1。

This is the input data set:这是输入数据集：

structure(list(campaign = c("campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x"), com_elm = c("campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_B1", "campaign_x_C3", 
"campaign_x_B1", "campaign_x_A1", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B1", "campaign_x_C3", "campaign_x_B1", "campaign_x_A1", 
"campaign_x_C3", "campaign_x_C3", "campaign_x_B1", "campaign_x_B2", 
"campaign_x_C3", "campaign_x_B1", "campaign_x_C3", "campaign_x_C3"
), com_elm_id = c(808001L, 811001L, 814001L, 509005L, 729060L, 
817002L, 820002L, 792002L, 793003L, 820003L, 824003L, 792002L, 
811001L, 787001L, 811001L, 468023L, 792002L, 812001L, 812001L, 
808001L, 811001L, 468023L, 468006L, 491014L, 825002L, 828002L, 
741001L, 825002L, 512001L, 733001L), recipient_id = c(54L, 54L, 
54L, 197L, 197L, 8388L, 8388L, 8426L, 8426L, 10903L, 10903L, 
14469L, 14469L, 17466L, 17466L, 17807L, 21666L, 23935L, 24287L, 
25412L, 25412L, 31361L, 31361L, 31361L, 31365L, 31365L, 40849L, 
40860L, 41737L, 41737L), step = c(3, 1, 2, 3, 3, 1, 2, 3, 3, 
1, 2, 3, 1, 3, 1, 1, 3, 1, 1, 3, 1, 1, 3, 3, 1, 2, 3, 1, 3, 3
), date = structure(c(19029, 19032, 19035, 18778, 18960, 19037, 
19040, 19016, 19019, 19040, 19043, 19015, 19032, 19011, 19032, 
18746, 19015, 19033, 19033, 19029, 19032, 18746, 18746, 18764, 
19044, 19047, 18969, 19044, 18781, 18962), class = "Date")), class = c("grouped_df", 
"tbl_df", "tbl", "data.frame"), row.names = c(NA, -30L), groups = structure(list(
    campaign = c("campaign_x", "campaign_x", "campaign_x", "campaign_x", 
    "campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
    "campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
    "campaign_x", "campaign_x", "campaign_x"), recipient_id = c(54L, 
    197L, 8388L, 8426L, 10903L, 14469L, 17466L, 17807L, 21666L, 
    23935L, 24287L, 25412L, 31361L, 31365L, 40849L, 40860L, 41737L
    ), .rows = structure(list(1:3, 4:5, 6:7, 8:9, 10:11, 12:13, 
        14:15, 16L, 17L, 18L, 19L, 20:21, 22:24, 25:26, 27L, 
        28L, 29:30), ptype = integer(0), class = c("vctrs_list_of", 
    "vctrs_vctr", "list"))), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -17L), .drop = TRUE))

This is the expected output dataframe:这是预期的 output dataframe：

structure(list(campaign = c("campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x"), com_elm = c("campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B2", "campaign_x_C3", "campaign_x_B1", "campaign_x_C3", 
"campaign_x_B1", "campaign_x_A1", "campaign_x_C3", "campaign_x_B1", 
"campaign_x_B1", "campaign_x_C3", "campaign_x_B1", "campaign_x_A1", 
"campaign_x_C3", "campaign_x_C3", "campaign_x_B1", "campaign_x_B2", 
"campaign_x_C3", "campaign_x_B1", "campaign_x_C3", "campaign_x_C3"
), com_elm_id = c(808001L, 811001L, 814001L, 509005L, 729060L, 
817002L, 820002L, 792002L, 793003L, 820003L, 824003L, 792002L, 
811001L, 787001L, 811001L, 468023L, 792002L, 812001L, 812001L, 
808001L, 811001L, 468023L, 468006L, 491014L, 825002L, 828002L, 
741001L, 825002L, 512001L, 733001L), recipient_id = c(54L, 54L, 
54L, 197L, 197L, 8388L, 8388L, 8426L, 8426L, 10903L, 10903L, 
14469L, 14469L, 17466L, 17466L, 17807L, 21666L, 23935L, 24287L, 
25412L, 25412L, 31361L, 31361L, 31361L, 31365L, 31365L, 40849L, 
40860L, 41737L, 41737L), step = c(3, 1, 2, 3, 3, 1, 2, 3, 3, 
1, 2, 3, 1, 3, 1, 1, 3, 1, 1, 3, 1, 1, 3, 3, 1, 2, 3, 1, 3, 3
), date = structure(c(19029, 19032, 19035, 18778, 18960, 19037, 
19040, 19016, 19019, 19040, 19043, 19015, 19032, 19011, 19032, 
18746, 19015, 19033, 19033, 19029, 19032, 18746, 18746, 18764, 
19044, 19047, 18969, 19044, 18781, 18962), class = "Date"), campaign_num_v1 = c(1, 
2, 2, 1, 2, 1, 1, 1, 2, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 2, 1, 
1, 2, 1, 1, 1, 1, 1, 2), campaign_num_v2 = c(1, 1, 1, 1, 2, 1, 
1, 1, 1, 1, 1, 1, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 
1, 1, 2)), class = c("grouped_df", "tbl_df", "tbl", "data.frame"
), row.names = c(NA, -30L), groups = structure(list(campaign = c("campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x", "campaign_x", "campaign_x", "campaign_x", "campaign_x", 
"campaign_x"), recipient_id = c(54L, 197L, 8388L, 8426L, 10903L, 
14469L, 17466L, 17807L, 21666L, 23935L, 24287L, 25412L, 31361L, 
31365L, 40849L, 40860L, 41737L), .rows = structure(list(1:3, 
    4:5, 6:7, 8:9, 10:11, 12:13, 14:15, 16L, 17L, 18L, 19L, 20:21, 
    22:24, 25:26, 27L, 28L, 29:30), ptype = integer(0), class = c("vctrs_list_of", 
"vctrs_vctr", "list"))), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -17L), .drop = TRUE))

This is an image of how first 10 observations of output dataframe would look like with the new campaign_num_v1 and campaign_num_v2 columns based on Approach 1 and Approach 2 described above added to the original dataframe:这是一张图像，显示了 output dataframe 的前 10 个观察值在基于上述方法 1和方法 2添加到原始 dataframe 的新campaign_num_v1和campaign_num_v2列时的样子：

输出数据框

I have tried multiple variations from solutions presented in different questions, but was not able to get it to work properly.我尝试了不同问题中提出的解决方案的多种变体，但无法使其正常工作。 Here is one of the simplest versions of coding I attempted, but does not work:这是我尝试过的最简单的编码版本之一，但不起作用：

  dat %>% group_by(campaign, com_elm, recipient_id) %>%
  arrange(recipient_id, date) %>%
  ungroup() %>%
  group_by(campaign, recipient_id ) %>%
  mutate(switch = ifelse(is.na(step > lead(step)), FALSE,step > lead(step))) %>%
  mutate(campaign_num_v1 = cumsum(step>lead(step)),
         campaign_num_v2 = cumsum(switch))

A solution using dplyr would be ideal, but I won't mind a different approach if there is no choice.使用 dplyr 的解决方案是理想的，但如果别无选择，我不介意采用其他方法。 Thanks.谢谢。

1 个解决方案

解决方案1
 2 已采纳  2022-04-04 21:54:49

You can use lag() with cumsum() , like this, and take care of the leading NA, using coalesce() thanks to Martin's helpful comment.您可以像这样将lag()与cumsum() ) 一起使用，并使用coalesce()处理领先的 NA，这要归功于 Martin 的有益评论。

df %>% arrange(campaign, recipient_id,date) %>% 
  group_by(campaign,recipient_id) %>% 
  mutate(campaign_num1 = cumsum(coalesce(lag(step)>=step, TRUE)),
         campaign_num2 = cumsum(coalesce(lag(step)>=step & date-lag(date)>14, TRUE))
)

Output: (first ten rows) Output：（前十行）

# A tibble: 30 x 8
# Groups:   campaign, recipient_id [17]
   campaign   com_elm       com_elm_id recipient_id  step date       campaign_num1 campaign_num2
   <chr>      <chr>              <int>        <int> <dbl> <date>             <int>         <int>
 1 campaign_x campaign_x_C3     808001           54     3 2022-02-06             1             1
 2 campaign_x campaign_x_B1     811001           54     1 2022-02-09             2             1
 3 campaign_x campaign_x_B2     814001           54     2 2022-02-12             2             1
 4 campaign_x campaign_x_C3     509005          197     3 2021-05-31             1             1
 5 campaign_x campaign_x_C3     729060          197     3 2021-11-29             2             2
 6 campaign_x campaign_x_B1     817002         8388     1 2022-02-14             1             1
 7 campaign_x campaign_x_B2     820002         8388     2 2022-02-17             1             1
 8 campaign_x campaign_x_C3     792002         8426     3 2022-01-24             1             1
 9 campaign_x campaign_x_C3     793003         8426     3 2022-01-27             2             1
10 campaign_x campaign_x_B1     820003        10903     1 2022-02-17             1             1

暂无
暂无

声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

相关问题 R，dplyr：如何根据group_by的大小来替换0值 - R,dplyr: How to replace 0 values based conditional on size of group_by R dplyr 根据 group_by 之后的条件应用 function - R dplyr apply a function based on a condition after group_by 根据条件在R dplyr中的group_by数据上插入新行 - Insert new row on group_by data in R dplyr based on condition R dplyr 条件变异与 group_by - R dplyr conditional mutate with group_by 如何使用dplyr基于R中的group_by对数据帧进行子集化 - How to use dplyr to subset dataframe based on group_by in R 在条件dplyr group_by中更正列标签 - Correct column label in conditional dplyr group_by 在 group_by 中过滤 + 突变不像在 group_by 中那样工作 + 在 dplyr R 中总结 - Filter in group_by + mutate not working as in group_by + summarise in dplyr R R: Group_by 取决于条件 - R: Group_by depending on condition R：找到与 dplyr 和 group_by 的绝对差异 - R: finding absolute difference with dplyr and group_by group_by grepl expression dplyr R. - group_by grepl expression dplyr R
相关标签
 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM