Sum by logical condition, in a dataset with groups - r

Question

I have a df like this:
I want to transform the continuous Age variable into a discrete one, that is equal a if the original was between 1 and 2, and b if it was betweem 3 and 4. Thus needing to aggregate the values of Value 1 and Value 2 by summing the entries associated with Age=1 + Age=2 and Age=3 + Age=4. The output would be something like this:
The 146 is the sum of the Value1 entry for Age=1 (75) and Age=2 (71).
I thought on using aggregate:
`df2 = df %>% group_by(Sex, Race) %>%
summarise(across(starts_with("Value"), fun))
Where fun would be some function that checks the Age values and sum accordingly. But i'm not much familiar with these dplyr functions and couldn't get it to work. Thanks for the help!
Data:
df = structure(list(Sex = c(1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2,
2, 2, 2), Race = c(1, 1, 1, 1, 2, 2, 2, 2, 1, 1, 1, 1, 2, 2,
2, 2), Age = c(1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4
), `Value 1` = c(75, 71, 52, 51, 24, 21, 70, 58, 67, 68, 36,
22, 91, 43, 33, 57), `Value 2` = c(22, 22, 49, 1, 20, 18, 34,
0, 27, 37, 31, 83, 29, 24, 10, 99)), row.names = c(NA, -16L), class = c("tbl_df",
"tbl", "data.frame"))

Answer 1

We can use case_when to do the recoding of 'Age' based on the values
library(dplyr)
df %>%
group_by(Sex, Race, Age = case_when(Age %in% 1:2 ~ 'a',
Age %in% 3:4 ~ 'b')) %>%
summarise(across(everything(), sum, na.rm = TRUE), .groups = 'drop')
-output
# A tibble: 8 x 5
# Sex Race Age `Value 1` `Value 2`
#* <dbl> <dbl> <chr> <dbl> <dbl>
#1 1 1 a 146 44
#2 1 1 b 103 50
#3 1 2 a 45 38
#4 1 2 b 128 34
#5 2 1 a 135 64
#6 2 1 b 58 114
#7 2 2 a 134 53
#8 2 2 b 90 109
Based on the OP's comment, if the original data have lots of categories, an easier option is cut or findInterval
df %>%
group_by(Sex, Race, Age = cut(Age, breaks = c(-Inf,
seq(0, 90, by = 5), Inf), labels = letters[1:20])) %>%
summarise(across(everything(), sum, na.rm = TRUE), .groups = 'drop')