I've got a loop in my code that I would like to rewrite so running the code takes a little less time to compete. I know you allways have to avoid loops in the code but I can't think of an another way to accomplice my goal.
So I've got a dataset "df_1531" containing a lot of data that I need to cut into pieces by using subset() (if anyone knows a better way, let me know ;) ). I've got a vector with 21 variable names on which I like assign a subset of df_1531. Furthermore the script contains 22 variables with constrains (shift_XY_time).
So, this is my code now...
# list containing different slots
shift_time_list<- c(startdate, shift_1m_time, shift_1a_time, shift_1n_time,
shift_2m_time, shift_2a_time, shift_2n_time,
shift_3m_time, shift_3a_time, shift_3n_time,
shift_4m_time, shift_4a_time, shift_4n_time,
shift_5m_time, shift_5a_time, shift_5n_time,
shift_6m_time, shift_6a_time, shift_6n_time,
shift_7m_time, shift_7a_time, shift_7n_time)
# List with subset names
shift_sub_list <- c("shift_1m_sub", "shift_1a_sub", "shift_1n_sub",
"shift_2m_sub", "shift_2a_sub", "shift_2n_sub",
"shift_3m_sub", "shift_3a_sub", "shift_3n_sub",
"shift_4m_sub", "shift_4a_sub", "shift_4n_sub",
"shift_5m_sub", "shift_5a_sub", "shift_5n_sub",
"shift_6m_sub", "shift_6a_sub", "shift_6n_sub",
"shift_7m_sub", "shift_7a_sub", "shift_7n_sub")
# The actual loop that I'd like to rewrite
for (i in 1:21) {
assign(shift_sub_list[i], subset(df_1531, df_1531$'PLS FFM' >= shift_time_list[i] & df_1531$'PLS FFM' < shift_time_list[i+1]))
}
Running the loop takes approximately 6 or 7 seconds. So, if anyone knows a better/cleaner or quicker way to write my code, I desperately like to hear your suggestion/opinion.
**Reproducible example **
mydata <- cars
dput(cars)
structure(list(speed = c(4, 4, 7, 7, 8, 9, 10, 10, 10, 11, 11,
12, 12, 12, 12, 13, 13, 13, 13, 14, 14, 14, 14, 15, 15, 15, 16,
16, 17, 17, 17, 18, 18, 18, 18, 19, 19, 19, 20, 20, 20, 20, 20,
22, 23, 24, 24, 24, 24, 25), dist = c(2, 10, 4, 22, 16, 10, 18,
26, 34, 17, 28, 14, 20, 24, 28, 26, 34, 34, 46, 26, 36, 60, 80,
20, 26, 54, 32, 40, 32, 40, 50, 42, 56, 76, 84, 36, 46, 68, 32,
48, 52, 56, 64, 66, 54, 70, 92, 93, 120, 85)), class = "data.frame", row.names = c(NA,
-50L))
dist_interval_list <- c( 0, 5, 10, 15,
20, 25, 30, 35,
40, 45, 50, 55,
60, 65, 70, 75,
80, 85, 90, 95,
100, 105, 110, 115, 120)
var_name_list <- c("var_name_1a", "var_name_1b", "var_name_1c", "var_name_1d",
"var_name_2a", "var_name_2b", "var_name_2c", "var_name_2d",
"var_name_3a", "var_name_3b", "var_name_3c", "var_name_3d",
"var_name_4a", "var_name_4b", "var_name_4c", "var_name_4d",
"var_name_5a", "var_name_5b", "var_name_5c", "var_name_5d",
"var_name_6a", "var_name_6b", "var_name_6c", "var_name_6d")
for (i in 1:24){
assign(var_name_list[i], subset(mydata,
mydata$dist >= dist_interval_list[i] &
mydata$dist < dist_interval_list[i+1]))
}
Starting with the 'reproducible' part and the information that the final aim is to summarize another column, it is possible to exploit the fact that the intervals are non-overlapping and simply use the cut function.
library(tidyverse)
mydata %>%
mutate(interval = cut(dist, breaks = dist_interval_list)) %>%
group_by(interval) %>%
summarise(sum = sum(speed))
This should be much faster and will also help you not to get lost in a messy environment full of variables (which are actually part of your data). You want to keep all your data in a single data frame as long as possible;) You probably want to follow with something like purrrlyr::invoke_rows at the final modeling step, if your function does not work with data frames.