使用R中特定列及其前一列的值检查条件

Question

I'm trying to create an output where a given condition is met: - the value in a given column must be 1 - the value in the previous column must be -1

As example, the input will be:

> cc <- matrix(c(1,-1,NA,-1,1,-1),  nrow=2, byrow = T)
> cc
     [,1] [,2] [,3]
[1,]    1   -1   NA
[2,]   -1    1   -1

I want to get an output with the same dimensions as my input and, in case the condition above is met, it will return "TRUE"; In case the condition is not met, the output should return FALSE or NA. As example:

> output<- matrix(c(NA,FALSE,NA,NA,TRUE,FALSE), nrow=2, byrow=T)
> output
     [,1]  [,2]  [,3]
[1,]   NA FALSE    NA
[2,]   NA  TRUE FALSE

The first column has NAs because we don't have a previous column but it can also be FALSE.

Answer 1

Here is a way to get the desired output:

> cbind(NA, cc[, -1, drop = F] == 1 & cc[, -ncol(cc), drop = F] == -1)
     [,1]  [,2]  [,3]
[1,]   NA FALSE    NA
[2,]   NA  TRUE FALSE

apply is not necessary. the drop = F is used to maintain the shape of the matrix in case it has only two columns.

Another way:

> x <- seq_len(ncol(cc))
> cc[, c(NA, x[-1])] == 1 & cc[, c(NA, head(x, -1))] == -1
     [,1]  [,2]  [,3]
[1,]   NA FALSE    NA
[2,]   NA  TRUE FALSE

Answer 2

Creating data frame

cc <- data.frame(matrix(c(1,-1,NA,-1,1,-1),  nrow=2, byrow = T))

cc
   X1  X2  X3
1   1  -1  NA
2  -1   1  -1

Looping through the data frame, changing row value based on value from previous column

# Loop starts from the last column
for(i in length(colnames(cc)):1){
    # Nested loop starts from first row
    for(j in 1:length(rownames(cc))){
        # Check if column is the first
        if(i == 1){
            cc[j,i] <- NA
        }
        # Check for NAs
        else if(is.na(cc[j,i]) | is.na(cc[j,i-1])){
            cc[j,i] <- NA
        }
        # Set to true if condition is met
        else if(cc[j,i] == 1 & cc[j,i-1] == -1){
            cc[j,i] <- TRUE
        } 
        # Else set to false
        else(cc[j,i] <- FALSE)
    }
}

This can obviously be further optimized, but I tried to keep the code as simple as possible.

Result

cc
   X1  X2  X3
1  NA   0  NA
2  NA   1   0

Explanation

I think you might have a logical error in your approach. This matrix cannot be looped through in sequential order (from column 1 to column n) and produce a reasonable result. If you start in column 1 and set it to NA (as it has no previous column) you will get

     [,1] [,2] [,3]
[1,]   NA   -1   NA
[2,]   NA    1   -1

In the next loop through column 2 you will inevitably get all NAs as the previous column is all NA (and not -1), and so forth.

     [,1] [,2] [,3]
[1,]   NA   NA   NA
[2,]   NA   NA   NA