Nested if statement and logic with strings in csv file in R - r

Question

I have a csv file that looks like this:
I try to create an algorithm that goes like this:
Iterate through each row;
If the condition is Success,
if T1 == P1, increase score one point
if T2 == P2, increase score one point
if T3 == P3, increase score one point
Else if the condition is Failure,
elif T1 != P1, increase score one point
elif T2 != P2, increase score one point
elif T3 != P3, increase score one point
However, I got stuck on 2 things:
When I say something like:
for (i in 1:4){
if (data[i,7] == "Success")
.......
There is a syntax problem because of using string with logic. How to get it right?
It doesn't calculate correctly when I state something like: if(data[i,1] == data[i,4]) {score = score+1}, but it does calculate correctly if I use numbers instead of letters in the csv file. Again, how to use strings with logic operators?
The other problem is using nested if statements. How to do it so I can use the algorithm above?
Thank you for your time!

Answer 1

We may also do this with across i.e. loop across the columns that starts_with 'T', then inside the loop, get the column names (cur_column()), replace the substring 'T', with 'P', and get its value, do a logical comparison, convert to numeric index by adding 1 (as R indexing starts from 1) to replace the values in vector (c(-1, 1)) based on the position index, and finally do a rowSums on the across output to create the 'total_score' column
library(dplyr)
library(stringr)
df %>%
mutate(total_score = rowSums(across(starts_with('T'),
~ c(-1, 1)[1 + (. == get(str_replace(cur_column(), 'T', 'P')))])))
-output
# A tibble: 4 x 5
T1 T2 P1 P2 total_score
<chr> <chr> <chr> <chr> <dbl>
1 a b a a 0
2 a a a a 2
3 a a a b 0
4 b a b b 0
data
df <- structure(list(T1 = c("a", "a", "a", "b"), T2 = c("b", "a", "a",
"a"), P1 = c("a", "a", "a", "b"), P2 = c("a", "a", "b", "b")), row.names = c(NA,
-4L), class = c("tbl_df", "tbl", "data.frame"))

Answer 2

a case_when structure can be used for your wish. Since we don't know how your data structure is, I created a dummy data which represents yours;
library(dplyr)
set.seed(1453)
scores <- data.frame(T1=sample(1:5,size = 200,replace = T),
T2=sample(1:5,size = 200,replace = T),
T3=sample(1:5,size = 200,replace = T),
P1=sample(1:5,size = 200,replace = T),
P2=sample(1:5,size = 200,replace = T),
P3=sample(1:5,size = 200,replace = T),
score=sample(50:100,size = 200,replace = T))
scores2 <- scores %>%
mutate(new_score=case_when(T1==P1 ~ score + 1,
T2==P2 ~ score + 1,
T3==P3 ~ score + 1,
TRUE ~ score - 1))
scores2%>%
head
Note: TRUE, means otherwise;
output;
T1 T2 T3 P1 P2 P3 score new_score
<int> <int> <int> <int> <int> <int> <int> <dbl>
1 4 2 2 1 5 5 64 63
2 3 5 4 2 1 3 82 81
3 5 1 5 4 5 5 89 90
4 2 5 3 4 5 1 62 63
5 3 5 4 3 2 4 53 54
6 3 1 4 1 3 2 82 81

Answer 3

If I got the problem right, each row is an observation, so I will compare each T column with the respective P column, than create a score for each comparison, finally I can sum them for each row.
Libraries
library(tidyverse)
Example Data
df <-
tibble(
T1 = c("a","a","a","b"),
T2 = c("b","a","a","a"),
P1 = c("a","a","a","b"),
P2 = c("a","a","b","b")
)
Code
df %>%
mutate(
S1 = if_else(T1 == P1, 1,-1),
S2 = if_else(T2 == P2, 1,-1)
) %>%
rowwise() %>%
mutate(total_score = sum(c_across(starts_with("S"))))
Output
# A tibble: 4 x 7
# Rowwise:
T1 T2 P1 P2 S1 S2 total_score
<chr> <chr> <chr> <chr> <dbl> <dbl> <dbl>
1 a b a a 1 -1 0
2 a a a a 1 1 2
3 a a a b 1 -1 0
4 b a b b 1 -1 0