R: Splitting a Dataset Into Parts Based on The Current Order

I am working with the R programming language.

I have the following dataset:

set.seed(123)
library(dplyr)

Patient_ID = 1:5000
gender <- c("Male","Female")
gender <- sample(gender, 5000, replace=TRUE, prob=c(0.45, 0.55))
gender <- as.factor(gender)


status <- c("Immigrant","Citizen")
status <- sample(status, 5000, replace=TRUE, prob=c(0.3, 0.7))
status  <- as.factor(status )

height = rnorm(5000, 150, 10)
weight = rnorm(5000, 90, 10)
hospital_visits = sample.int(20,  5000, replace = TRUE)

disease = sample(c(TRUE, FALSE), 5000, replace = TRUE)

my_data = data.frame(Patient_ID, gender, status, height, weight, hospital_visits, disease)

My Question: Suppose I select all rows with male immigrants - I want to split this resulting dataset into 4 parts such that the top 25% tallest are in one dataset, the second 25% tallest are in another dataset, etc.

Currently, I am doing this manually:

part_1 = my_data[my_data$gender == "Male" & my_data$status == "Immigrant",]
part_1 = part_1 %>% arrange(desc(height))

limits = as.integer(seq(1, nrow(part_1), by = 0.25*nrow(part_1)))
limits = c(limits, nrow(part_1))

> limits
[1]   1 163 326 489 651

h_1 = part_1[1:163,]
h_2 = part_1[164:326,]
h_3 = part_1[327:489,]
h_4 = part_1[490:651,]

I can verify these results manually by making sure that there is no overlap in heights:

> range(h_1$height)
[1] 157.3934 182.7591
> range(h_2$height)
[1] 149.8167 157.3084
> range(h_3$height)
[1] 143.8353 149.7927
> range(h_4$height)
[1] 111.5468 143.8141

My Problem: I would now like to do this using a function/loop (i.e. in case I want to make more than 4 splits). But when I try doing this using a function/loop, the splits are not being made properly (i.e. note the overlapping ranges):

create_h <- function(part_1) {
    limits = as.integer(seq(1, nrow(part_1), by = 0.25*nrow(part_1)))
    limits = c(limits, nrow(part_1))
    h_list = list()
    for (i in 1:(length(limits)-1)) {
        h_list[[i]] = part_1[(limits[i]+1):limits[i+1],]
    }
    return(h_list)
}

for (i in seq_along(h_list)) {
    assign(paste0("h_", i), h_list[[i]])
}

>  range(h_1$height)
[1] 120.7571 182.7591
>  range(h_2$height)
[1] 115.3251 178.7666
>  range(h_3$height)
[1] 125.9139 173.9946
> range(h_4$height)
[1] 124.3712 173.7773

Can someone please show me what I am doing wrong and what I can do to fix this?

Thanks!

I can't reproduce your problem. You don't show how you call create_h, but I assume it's something like this. We can then verify the ranges (easier to do in the list than after assigning)

h_list = create_h(part_1)
# [[1]]
# [1] 157.3934 178.7666
# 
# [[2]]
# [1] 149.8167 157.3084
# 
# [[3]]
# [1] 143.8353 149.7927
# 
# [[4]]
# [1] 111.5468 143.8141

My best guess as to the problem is that you have an old version of h_list in your environment and you never updated it after debugging your function.

You do have one bug, the first row is omitted, even when i = 1 you have h_list[[i]] = part_1[(limits[i]+1):limits[i+1],], and (limits[i]+1) will start at 2.

Instead of this approach, we can use the convenience function dplyr::ntile to simplify everything:

n_quant = 4
h_list2 = part_1 |>
  group_by(n_tile = ntile(n = n_quant)) |>
  group_split()

## verify
lapply(h_list2, \(x) c(n = nrow(x), range = range(x$height)))
# [[1]]
#        n   range1   range2 
# 163.0000 157.3934 182.7591 
# 
# [[2]]
#        n   range1   range2 
# 163.0000 149.8167 157.3084 
# 
# [[3]]
#        n   range1   range2 
# 163.0000 143.8353 149.7927 
# 
# [[4]]
#        n   range1   range2 
# 162.0000 111.5468 143.8141