英文:
Filtering longitudinal data by last recorded event, and counting number of rows that meet criteria (and storing that value)
问题
Here's the translated code portion:
我有一些（医疗）纵向数据，用于检测（或未检测到）支气管扩张症。我希望统计最新随访/检查时有无支气管扩张症的参与者人数。然后，我希望存储有支气管扩张症和没有支气管扩张症的患者人数。这样，我可以轻松地将这个数字/值插入到医学报告中（我还有其他需要这种方法的结果）。我正在寻找是否有更简单/更准确的方法来完成我正在做的事情。示例数据框：| id | 随访日期 | 支气管扩张症 || --- | --- | --- || 1 | 2022年01月01日 | 否 || 1 | 2022年01月12日 | 是 || 2 | 2022年02月07日 | 是 || 2 | 2023年05月06日 | 是 || 3 | 2021年01月05日 | 否 || 3 | 2022年03月12日 | 否 |这是我的当前代码：```Rdf1 <- df %>%group_by(id) %>%filter(date_of_followup == max(date_of_followup)) %>%filter(bronchiectasis == "是")bronchiectasisyes <- sum(df1$bronchiectasis == "是")df2 <- df %>%group_by(id) %>%filter(date_of_followup == max(date_of_followup)) %>%filter(bronchiectasis == "否")bronchiectasisno <- sum(df2$bronchiectasis == "否")
这段代码可以获得我想要的结果，但可能有点混乱，我怀疑是否有更优雅/更健壮的方法。
<details><summary>英文:</summary>I have some (medical) longitudinal data on the detection (or not) of bronchiectasis. I wish to count the number of participants with and without bronchiectasis at their latest follow up/check. I then wish to store the number of patients who do have bronchiectasis and the number who don't. This is so I can then insert that number/value into a qmd report easily. (I have other outcomes where I need this method).Is there a simpler/more accurate method of what I'm doingExample df| id | date_of_followup | bronchiectasis || --- | --- | --- || 1 | 01/01/2022 | N || 1 | 01/12/2022 | Y || 2 | 02/07/2022 | Y || 2 | 5/6/2023 | Y || 3 | 01/05/2021 | N || 3 | 3/12/2022 | N |This is my current code
df1 <- df %>%
group_by(id) %>%
filter(date_of_followup == max(date_of_followup )) %>%
filter(bronchiectasis=="Y")
bronchiectasisyes <- sum(df1$bronchiectasis == "Y")
df2 <- df%>%
group_by(id) %>%
filter(date_of_followup == max(date_of_followup )) %>%
filter(bronchiectasis=="N")
bronchiectasisno <- sum(df2$bronchiectasis == "N")
This gets me the desired outcome, but perhaps is a bit messy and I suspect there is a more elegant/robust method</details># 答案1**得分**: 0以下是翻译好的部分：修复日期，按最大id筛选，然后进行汇总：```rlibrary(dplyr)df %>%mutate(date_of_followup = as.Date(date_of_followup, format="%m/%d/%Y")) %>%slice_max(date_of_followup, by = id) %>%count(bronchiectasis)# bronchiectasis n# 1 N 1# 2 Y 2
要确认这是否有效，请查看中间步骤：
df %>%mutate(date_of_followup = as.Date(date_of_followup, format="%m/%d/%Y")) %>%slice_max(date_of_followup, by = id)# id date_of_followup bronchiectasis# 1 1 2022-01-12 Y# 2 2 2023-05-06 Y# 3 3 2022-03-12 N
（我假设您的日期使用m/d/Y格式...如有需要请进行修复。）
英文:
Fixing the date, filtering for max-per-id, then summarizing:
library(dplyr)df %>%mutate(date_of_followup = as.Date(date_of_followup, format="%m/%d/%Y")) %>%slice_max(date_of_followup, by = id) %>%count(bronchiectasis)# bronchiectasis n# 1 N 1# 2 Y 2
To confirm this is working, see the interim step:
df %>%mutate(date_of_followup = as.Date(date_of_followup, format="%m/%d/%Y")) %>%slice_max(date_of_followup, by = id)# id date_of_followup bronchiectasis# 1 1 2022-01-12 Y# 2 2 2023-05-06 Y# 3 3 2022-03-12 N
(I'm assuming m/d/Y format for your dates ... fix as appropriate.)
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论