R for loop for filtering and summarizing dataframe (with dyplr)?-CodePudding

I am using a simple command with dyplr to first filter a dataframe by two columns and then report the sum of another column. However I would like to create a loop so that the filtering criteria can be automated by a list of values. For example the code for a single instance:

library(dplyr)
df = data.frame(Category1 = sample(c("FilterMe","DoNotFilterMe"), 15, replace=TRUE), 
          Category2 = sample(c("1","3","5","10"),15, replace=TRUE),
          Value = 1:15)

df %>%
filter(Category1=="FilterMe" & Category2="1") %>%
summarize(result=sum(Value))

This works perfectly and I get a single value of 15. However I would like to loop the command such that I can do multiple values for Category2 defined by a list of integers (not sequential). I want it to loop for each value of i and provide a different output value each time. I tried the code below but was left with a null value.

library(dplyr)
for (i in c(1,3,5,10){
df %>%
filter(Category1=="FilterMe" & Category2="i") %>%
summarize(result=sum(Value))}

If there is another way besides loop that would fulfill the same objective that is fine by me.

CodePudding user response：

If I understood what you want to do, you are looking for group_by.

library(dplyr)
df %>%
   filter(Category1 =="FilterMe") %>%
   group_by(Category2) %>%
   summarize(result=sum(Value))

CodePudding user response：

We don't need a loop. It can be simplified with %in% instead of == and then do group_by sum approach

library(dplyr)
df %>%
  filter(Category1=="FilterMe" & Category2 %in% c(1, 3, 5, 10)) %>%
  group_by(Category2) %>%
  summarize(result=sum(Value))

-output

# A tibble: 4 × 2
  Category2 result
  <chr>      <int>
1 1              4
2 10            15
3 3             17
4 5             19

With a for loop, we need to store the output in each of the iteration i.e. a list

v1  <- c(1, 3, 5, 10)
lst1 <- vector('list', length(v1))
for (i in seq_along(v1)){
  lst1[[i]] <- df %>%
      filter(Category1=="FilterMe" & Category2 ==v1[i]) %>%
      summarize(result=sum(Value))

}

-output

> lst1
[[1]]
  result
1      4

[[2]]
  result
1     17

[[3]]
  result
1     19

[[4]]
  result
1     15

Or may directly store the output in a list with map/lapply

library(purrr)
map(c(1, 3, 5, 10), ~ 
       df %>%
         filter(Category1 == "FilterMe", Category2 == .x) %>%
         summarise(result = sum(Value)))

-output

[[1]]
  result
1      4

[[2]]
  result
1     17

[[3]]
  result
1     19

[[4]]
  result
1     15