Searching through a matrix in a specific pattern

Question

I trying to do some fetching values in R. I am not an experienced programming.

I have a function to search the matrix in an specific manner. This is the matrix:

k2=c(3,6,4,NA,NA,NA,NA)
k3=c(2,7,5,NA,NA,NA,NA)
k4=c(7,9,5,2,1,8,12)

df2=data.frame(k2,k3,k4)
m2=as.matrix(df2)
m2

     k2 k3 k4
[1,]  3  2  7
[2,]  6  7  9
[3,]  4  5  5
[4,]  NA NA  2
[5,]  NA NA  1
[6,]  NA NA  8
[7,]  NA NA 12

The main idea to find a value in the k4 row. If the value is not the same for that row, the function should check the values in the same row, and go to these rows instead to find the value.

So if the value is less than val in [1,k4] then it should take the second column value in the same row ([1,k3])as the row number to go for and check the value in the third column in that row [value of row,3]. If the value is the greater, itshould do same thing as before, but with [1,k2]. And if the value is again not the needed one it should again check and repeat the same procedure for that next row until it gets the match.

Example :

So if I need val=5 then it should check if it is the same value as [1,k4] which is 7 and it is not, it should then go to third low check [3,k4] and as you can see it is the same, which means it is going to print 5. If I need the value I need is 12, then if should the logic be like this : the val is not the same as 7, so go to the third row, check with 9, oh its not the same value, the I will go to the 7 row, and there it is, its same val, I will print then...

My code so far/My progress

I know that I have to search the third every column time, so that means, beginning with the first

for (i in df2[1,3])

And then check if each value in that column is not equal to val:

if (!i == val)

And here comes the tricky part for me. I understand that I have to make a check if its less or greater, which means that is should somehow be like this

  if (!i == val) 
   #then check if its less, if yes, go that row
   #if it greater then go then go that row
   # go to that row and check the same thing there

And I do not know how to do that, so that it repeats itself for every row until it finds the value in val. I know that I can assing the row for smaller value as n-2 and the row for the bigger value as n-1, but I am strugling here to make it work. As I said, I have very little experience in programming.

Thanks a lot!

I really need help with this.

Answer 1

I am going to re-state your problem so that we can be sure we're together.

We have a value to find, and a matrix to look through. The matrix is like a map: it gives directions. We start on the first row, and compare to the k4 column.

• If the value is equal to the k4 column on this row, we are done, success!
• If the value is less than the k4 column on this row, go to the row number indicated by the k2 column.
• If the value is greater than the k4 column, go to the row number indicated by the k3 column.
• Continue until there is a match.

As you say, it must "repeats itself for every row until it finds the value". Since we do not know how many repetitions, we must use a while loop, not a for loop. A while loop will keep trying until it is done. (This is dangerous because if it does not find an answer it will keep trying forever! So maybe we will put in a condition to stop if it tries too long.)

k2=c(3,6,4,NA,NA,NA,NA)
k3=c(2,7,5,NA,NA,NA,NA)
k4=c(7,9,5,2,1,8,12)

m2 = cbind(k2, k3, k4)

my_search = function(value, matrix) {
    current.row = 1
    try.counter = 1
    while(value != matrix[current.row, 3]) {
        if (value < matrix[current.row, 3]) {
            current.row = matrix[current.row, 1]
        }
        if (value > matrix[current.row, 3]) {
            current.row = matrix[current.row, 2]
        }

        # This is the condition to stop if we can't find an answer
        # To keep the algorithm from getting caught in a circle.
        if (try.counter > 1e6) stop("Couldn't find an answer :(")
        try.counter = try.counter + 1
    }
    cat(paste("The successful row number is", current.row, "\n"))
    cat("It's values are:\n")
    print(matrix[current.row, ])
}

Usage examples:

> my_search(12, m2)
k2 k3 k4 
NA NA 12 
> my_search(5, m2)
k2 k3 k4 
 4  5  5 

If you want to be able to assign the output of the function, you should add another line to the function specifying what you want to return, eg return(matrix[current.row, ]) .