Loop through the columns to search for multiple variables in R - r

Question

Sorry, that this is a follow-up question. I am trying to count how many 'S' and 'T' appears in each column as 'downstream' from 1 to 10 rows and then as 'upstream' from 15 to 25.
ST <- data.frame(scale = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0),
aa = c('A','C','D','E','F','G','H','I','K','L','M','N','P','Q','R','S','T','V','W','Y'))
#input (example)
V1 V2 V3 V4 V5
1 C D E R N
2 C A M K P
3 V T Q Q E
4 A T S S S
5 C D E R N
6 C A M K P
7 V T Q Q E
8 A T S S S
9 R V D S A
10 W R H I C
11 S N I P T
12 Q A S D E
13 C D E R N
14 C A M K P
15 V T Q Q E
16 A T S S S
17 C D E R N
18 C A M K P
19 V T Q Q E
20 A T S S S
21 R V D S A
22 W R H I C
23 S N I P T
24 G A D S S
25 N T T S A
When I had a data frame with 'S' only, the script below worked but with 'ST', it doesn't. Could someone tell me why? Of course, I could get 'S' and 'T' separately and then add it later but is there a way to do it through this single data frame 'ST'?
#sum values from positions 1 to 10 and then from 15 to 25 works well for 1 letter only
count_aa <- df_trial %>%
summarise(across(everything(), ~ c(sum(.[1:10] == 'T'), sum(.[15:25] == 'T')))) %>%
mutate(categ = c('upstream', 'downstream'), .before = 1)
#view(count_aa)
df_count_aa<- as.data.frame(t(count_aa))
#view(df_count_aa)

Answer 1

We can use %in% instead of == when there are more than one element to compare
library(dplyr)
df_trial %>%
summarise(across(everything(), ~
c(sum(.[1:10] %in% c('S', 'T')),
sum(.[15:25] %in% c('S', 'T'))))) %>%
mutate(categ = c('upstream', 'downstream'), .before = 1)
-output
# categ V1 V2 V3 V4 V5
#1 upstream 0 4 2 3 2
#2 downstream 1 5 3 5 4
The == is doing elementwise comparison. If we do the == with more than one element as == c("S", "T"), then it does a recycling of the vector elements to the entire length of the column resulting i.e. 'S' gets compared to the first element of the colum, 'T' to second element, 'S' again to 3rd element and so on... i.e. the comparison would be based on position
In base R we can do colSums
colSums(df_trial == 'S') + colSums(df_trial == 'T')

Answer 2

In base R, you can do this sapply :
data.frame(categ = c('upstream', 'downstream'),
sapply(df_trial, function(x)
c(sum(x[1:10] %in% c('S', 'T')), sum(x[15:25] %in% c('S', 'T')))))
# categ V1 V2 V3 V4 V5
#1 upstream 0 4 2 3 2
#2 downstream 1 5 3 5 4

Answer 3

Using base R
> rbind(downstream = sapply(df[1:10,], function(x) sum(grepl('[ST]',x))),
+ upstream = sapply(df[15:25,], function(x) sum(grepl('[ST]',x))))
V1 V2 V3 V4 V5
downstream 0 4 2 3 2
upstream 1 5 3 5 4
>
Data Used:
> dput(df)
structure(list(V1 = c("C", "C", "V", "A", "C", "C", "V", "A",
"R", "W", "S", "Q", "C", "C", "V", "A", "C", "C", "V", "A", "R",
"W", "S", "G", "N"), V2 = c("D", "A", "T", "T", "D", "A", "T",
"T", "V", "R", "N", "A", "D", "A", "T", "T", "D", "A", "T", "T",
"V", "R", "N", "A", "T"), V3 = c("E", "M", "Q", "S", "E", "M",
"Q", "S", "D", "H", "I", "S", "E", "M", "Q", "S", "E", "M", "Q",
"S", "D", "H", "I", "D", "T"), V4 = c("R", "K", "Q", "S", "R",
"K", "Q", "S", "S", "I", "P", "D", "R", "K", "Q", "S", "R", "K",
"Q", "S", "S", "I", "P", "S", "S"), V5 = c("N", "P", "E", "S",
"N", "P", "E", "S", "A", "C", "T", "E", "N", "P", "E", "S", "N",
"P", "E", "S", "A", "C", "T", "S", "A")), row.names = c(NA, -25L
), class = c("tbl_df", "tbl", "data.frame"))
>