How to arrange and select the following row in R?-CodePudding

Let's say I have a dataset that looks like the following, and then, I proceed to group the rows by arranging them using NU_DN and DATE_HOSP:

  NU_DN <- c("55284765", "55293825" , "55284765", "55293825", "55318351", "55293825")
    DATE_HOSP <- c("2012-05-08", "2012-06-17", "2012-09-13", "2012-01-13", "2012-01-10", "2012-05-04" )
    MOTIV <- c("2.1 - Reason 1", "1.2 - Reason 3", "6.7 - Reason 8", "2.2 - Reason 2", "3.3 - Reason 4", "6.9 - Reason 10")
    
    data <- as.data.frame(cbind(NU_DN, DATE_HOSP, MOTIV))

data <- data %>%
  arrange(NU_DN, DATE_HOSP)%>%
  type.convert(as.is = TRUE)%>%
  group_by(NU_DN) %>%
  ungroup()

After that's done, I get the following dataframe:

 NU_DN       DATE_HOSP         MOTIV
55284765    2012-05-08    2.1 - Reason 1
55284765    2012-09-13    6.7 - Reason 8
55293825    2012-01-13    2.2 - Reason 2
55293825    2012-05-04    6.9 - Reason 10
55293825    2012-06-17    1.2 - Reason 3
55318351    2012-01-10    3.3 - Reason 4

What I want to do is to select the following row of the cluster if "MOTIV" starts with either 2 or 3. So if MOTIV for for NU_DN 55293825 is "2.2 - Reason 2", I want to select the next row that it's MOTIV is '6.9 - Reason 10" AND "1.2 - Reason 3". My desired output is the following:

 NU_DN       DATE_HOSP         MOTIV
55284765    2012-09-13    6.7 - Reason 8
55293825    2012-05-04    6.9 - Reason 10
55293825    2012-06-17    1.2 - Reason 3

Is there any way I can do that?

CodePudding user response：

How about removing the lines that start with either 2 or 3?

str_detect() allows detecting the presence of strings that start with 2 (^2) or 3 (^3).

negate = TRUE returns non-matching elements.

library(tidyverse)

data %>% 
  dplyr::filter(str_detect(MOTIV, c("^2", "^3"), negate = TRUE))

CodePudding user response：

If you want all non-matched rows:

data %>% filter(!grepl("^3|^2", MOTIV))

1 55284765 2012-09-13 6.7 - Reason 8 
2 55293825 2012-05-04 6.9 - Reason 10
3 55293825 2012-06-17 1.2 - Reason 3

If you only want the one row that follows any match (it assumes it will also not be a match):

idx <- data %>%  
  mutate(i=grepl("^3|^2", MOTIV)) %>%
  pull(i)

data[which(idx) 1,] %>%  
  drop_na()

1 55284765 2012-09-13 6.7 - Reason 8 
2 55293825 2012-05-04 6.9 - Reason 10