使用dplyr按就诊序列以0.01步长扩展数据框的实现方法
实现代码
tidyverse 版本
library(tidyverse)
df %>% separate_rows(visit_str, sep = ",") %>% group_by(patient) %>% mutate( y_new = y_start + (row_number() - 1) * 0.01, visit_str = if_else(row_number() != n(), paste0(visit_str, ","), visit_str) ) %>% ungroup() %>% select(patient, y_new, visit_str)
输出效果:
# A tibble: 19 × 3 patient y_new visit_str <chr> <dbl> <chr> 1 k2-01-003 0.292 "9," 2 k2-01-003 0.302 "10," 3 k2-01-003 0.312 "11," 4 k2-01-003 0.322 "12," 5 k2-01-003 0.332 "13," 6 k2-01-003 0.342 "14," 7 k2-01-003 0.352 "15," 8 k2-01-003 0.362 "16," 9 k2-01-003 0.372 "17," 10 k2-01-003 0.382 "18" 11 k2-01-008 0.048 "9," 12 k2-01-008 0.058 "10," 13 k2-01-008 0.068 "11," 14 k2-01-008 0.078 "12," 15 k2-01-008 0.088 "13," 16 k2-01-008 0.098 "14," 17 k2-01-008 0.108 "15," 18 k2-01-008 0.118 "16," 19 k2-01-008 0.128 "17"
Base R 版本
无额外包依赖,直接运行:
res <- do.call(rbind, lapply(seq_len(nrow(df)), function(i) { visits <- unlist(strsplit(df$visit_str[i], ",")) y_vals <- df$y_start[i] + (seq_along(visits) - 1) * 0.01 data.frame( patient = df$patient[i], y_new = y_vals, visit_str = ifelse(seq_along(visits) != length(visits), paste0(visits, ","), visits) ) }))
注意点
- 步长值可直接修改代码中
0.01的参数适配不同业务规则 - 若需保留原表
nof_visits字段,可在tidyverse管道的select()步骤中添加该列,base R方法在构造data.frame时加入对应字段即可 - 代码自动匹配示例中“每个患者除最后一个就诊外,其余就诊后追加逗号”的格式要求
内容的提问来源于stack exchange,提问作者littleworth
相关产品推荐
相关产品推荐

