For each row, count number of other rows that fall within a time interval

example_data.csv

example_for_loop_advantage.R

# Load libraries
library(tidyverse) # CRAN v1.3.1
library(lubridate) # CRAN v1.7.10


# Read in example data
in_data <- read_csv("example_data.csv")


# Modify to create intervals and one year prior start_date
df <- in_data %>%
  mutate(
    minus1Y = start_date %m-% years(1),
    interval = interval(ymd(minus1Y), ymd(start_date)),
    count = 0
  )


# Put back results into this data frame
df_out <- data.frame()


# Loop through rows and compare each row to the rest of the data frame
for (i in 1:nrow(df)) {
  # Pull out row of interest
  tmp <- df[i,]

  # Loop through other intervals in the data frame
  # Count number of rows that have a stop_date within the interest interval
  n_count <- df %>%
    dplyr::setdiff(tmp) %>%
    mutate(in_interval = stop_date %within% tmp$interval) %>%
    filter(in_interval) %>%
    nrow()

  # Save back information into data
  tmp$count <- n_count
  df_out <- bind_rows(df_out, tmp)
}

df_out