R: 'If' condition not working correctly when inside this 'for' loop

Question

So, i have this small database that i'm working with in RStudio, and I have this situation where i need to classify some students based on which classes they flunked.

for (i in 1:(nrow(BDbruno2)-1))
{
  #avoiding exploitations, i tested without it but my problem still continues
  if(i >= nrow(BDbruno2))
  {
    break 
  }else 
  {
    #checking is the codes are still the same
    if((BDbruno2$Code[i] == BDbruno2$Code[i+1]) && (i < nrow(BDbruno2)))
    {
      auxIndice <- BDbruno2$nLinha[i]
      auxTurmas <- BDbruno2$tempo[i]

      for(j in (i+1):nrow(BDbruno2))
      {
        #checking if codes are the same, if FALSE, i save all classes and save in a string for all codes
        if(BDbruno2$Code[j-1] != BDbruno2$Code[j]){
          BDbruno2$turmasCalc1[auxIndice] <- paste0(auxTurmas, collapse = " ")
          #skipping the same codes that i already checked
          i <- j
          #i tested without this break, only makes my code to take longer to finish
          break
        } else
        {
          #saving all rows where the code is the same
          auxIndice <- c(auxIndice, BDbruno2$nLinha[j])
          #this line below is where i get my problem:
          #it receives the classes from the same code, but when going further in the loop, this var gets messed
          auxTurmas <- c(auxTurmas, BDbruno2$tempo[j])
        }
      }
    }
  }
}

The auxTurmas var won't return the expected result, but it does when i run the code line-by-line.

There are these 4 cases where i get the same student (different rows), i save all his classes in a new var (turmasCalc1), which are 1 4 7 8, and all his rows gets theses numbers, but when inside a for-loop, the first one is correct, the second fails (4 7 8), the other two also fail(7 8).

Weirdly, if i run it from i = nLinha in 1:46, it works as it should, but i need it for all my cases (which happens a lot). I'm not sure, but it doesn't seem to be a issue with the break i used, but i can't see what makes this strange thing happen. Can somebody give me a light?

Edit: Sorry for the lack of information, here is a sample of the data frame, it's supposed to return 1 4 7 8 on row nLinha = 46:49, similar problem occurs other times on same data frame.

Code             Calculo1_Turma          tempo   turmasCalc1  nLinha
1635340632       2014/1 - MAT154-B          11         11     45
1638717605       2009/1 - MAT154-E          1     1 4 7 8     46
1638717605       2010/3 - MAT154-I          4       4 7 8     47
1638717605       2012/1 - MAT154-A          7         7 8     48
1638717605       2012/3 - MAT154-D          8         7 8     49
1643222643       2011/1 - MAT154-D          5         5 6     50
1643222643       2011/3 - MAT154-B          6         5 6     51
1645485641       2009/1 - MAT154-B          1           1     52

This is what I'm trying to get:

Code             Calculo1_Turma          tempo   turmasCalc1  nLinha
1635340632       2014/1 - MAT154-B          11         11     45
1638717605       2009/1 - MAT154-E          1     1 4 7 8     46
1638717605       2010/3 - MAT154-I          4     1 4 7 8     47
1638717605       2012/1 - MAT154-A          7     1 4 7 8     48
1638717605       2012/3 - MAT154-D          8     1 4 7 8     49
1643222643       2011/1 - MAT154-D          5         5 6     50
1643222643       2011/3 - MAT154-B          6         5 6     51
1645485641       2009/1 - MAT154-B          1           1     52

Edit2: Sorry again for the confusion. Here is the dpyr-generated code:

BDbruno2 <- structure(list(Code = c("1634171640", "1634171640", "1634171640", "1635340632", "1638717605", "1638717605", "1638717605", "1638717605", "1643222643", "1643222643", "1645485641"), Calculo1_Turma = c("2009/1 - MAT154-D", "2009/3 - MAT154-A", "2010/3 - MAT154-I", "2014/1 - MAT154-B", "2009/1 - MAT154-E", "2010/3 - MAT154-I", "2012/1 - MAT154-A", "2012/3 - MAT154-D", "2011/1 - MAT154-D", "2011/3 - MAT154-B", "2009/1 - MAT154-B"), tempo = c(1, 2, 4, 11, 1, 4, 7, 8, 5, 6, 1), turmasCalc1 = c("1", "2", "4", "11", "1 4 7 8", "4 7 8", "7 8", "7 8", "5 6", "5 6", "1"), nLinha = 42:52), .Names = c("Code", "Calculo1_Turma", "tempo", "turmasCalc1", "nLinha"), row.names = c(162L, 305L, 714L, 3880L, 210L, 715L, 887L, 924L, 2157L, 2446L, 60L), class = "data.frame")

This one generates a few more row below, which are working as they should. Just a recap: i'm supposed to get 1 4 7 8 on turmasCalc1 where nLinha in 46:49, but there seems to be a issue with the i index. This issue happens when the same code appears 3 or more times, not only specifically 4.

Answer 1

I'm going to take a stab at starting this. I'm probably not going to get your output right (because I really have no idea what you want) - but I'm hoping this gets you started in the right direction.

library(dplyr)
BDbruno2 %>%
  group_by(CODIGO) %>%
  summarize(turmasCalc1 = paste(tempo, collapse = " ")) %>%
  left_join(select(BDbruno2, -turmasCalc1), .)

This is an R-style solution to the problem. R is an interpreted language - which is great for changing things frequently on the fly, but fairly slow for heavy computation. The solution is what is known in R as vectorized functions - which means that the heavy computation stuff and loops are implemented in a compiled language, and R just passes the data on. This means that writing loops IN R is almost always a bad idea - we want to call the vectorized version of the function which is going to do the looping for us. Here's a longer version of what I just said with some examples: http://alyssafrazee.com/2014/01/29/vectorization.html

More recently, another revolution has happened in R - which is manifested in a suite of packages called tidyverse, one of which I used above - dplyr. The entire suite of packages is amazing, but the big one is the use of pipes (%>%). All they do is take the result of the previous function, and set it as the first argument of the next function - but they allow us to linearize the function calls, and see what is going on.

Take the code above - I'm first grouping by CODIGO (which I'm assuming is the same as Code in the for loop you provided). No more looking for whether the code is the same or not, we are looking at the data in chunks, and the code is the same for everything in the chunk. The next function is summarize - which says that we want to generate a single summary for each code, and we're going to get it by pasting together the elements of tempo.

Finally, we're going to merge this back in with the original data set using left_join. Here, I want to make turmasCalc1 the last variable instead of the first, so I wanted the original data set (DBbruno2 to be the first arg). This is why it's called with a single dot after the column - I'm overriding the default behavior of entering in the result as the first argument, and making it the second instead.