Question

Here is the data:

   # vector1 dates 
     reading.dates <- as.Date(c("2012-02-13","2012-02-20","2012-02-28"))

     mydat <- data.frame (ID = c("A", "B", "C", "D"), Date1 = c(1:4),
      Date2 = c(5:8), Date3 = c(11:14))

     mydat
  ID Date1 Date2 Date3
1  A     1     5    11
2  B     2     6    12
3  C     3     7    13
4  D     4     8    14

Here is what I want to do:

for each ID levels,

Between date1 and date2

calculate difference is dates –

reading.dates[2] - reading.dates[1] = 7 = say "Y"

Add correponding date values and average –

                (date1 + date2) / 2 = say "X"
                   then multiply = X * Y


 for indiviudual A:((1 + 5) / 2)* 7 )

Between date2 and date3

Similarly between date2 and date3 ….so on to end of the file, infact I have more variables

reading.dates[3] – reading.dates[2]
Time difference of 8 days

for A indvidual (5+11)/2) * 8

then add the all values.

for A it would be

((1 + 5) / 2)* 7 ) + ((5+11)/2) * 8) 

Thanks

Answer 1

Try this code, I think it should work for any size of data.frame and date vector:

calculateData <- function(dates, data)
{
    if (length(dates) != (ncol(data)-1))
        stop("The columns and dates doesn't match")

    out <- as.data.frame(NULL)
    for (i in 1:nrow(data))
    {
        x <- NULL
        for (j in 2:length(data))
            x <- c(x, (data[i,j]+data[i,j+1])/2 * difftime(dates[j], dates[j-1], "days"))
        out <- rbind(out, c(data[i,1], sum(x)))
    }

    colnames(out) <- c("ID", "VAL")
    return(out);
}

reading.dates <- as.Date(c("2012-02-13","2012-02-20","2012-02-28"))
mydat <- data.frame (ID = c("A", "B", "C", "D"), Date1 = c(1:4), Date2 = c(5:8), Date3 = c(11:14))

cd <- calculateData(reading.dates, mydat)
print(cd)

You can change yourself how you want to have the data returned.

   > print(cd)
      ID VAL
    1  1  85
    2  2 100
    3  3 115
    4  4 130