Functions for finding first and last variables in the vector

Question

I am looking for an R function to find first and last variables in the vector similar to min for minimal and max for maximal. I know I can calculate the length of the vector, and go from there but it does not work for what I need it for.

I have a following dataset (in reality much larger):

a<-(c("2013-02-25","2013-03-13","2013-04-24","2013-05-12","2013-07-12","2013-08-11","actual_exam_date"))
b<-c(300,230,400,NA,NA,NA,"2013-04-30")
c<-c(NA,260,410,420,NA,NA,"2013-05-30")
d<-c(300,230,400,NA,370,390,"2013-08-30")
df<-as.data.frame(rbind(b,c,d))
colnames(df)<-a
rownames(df)<-(c("student 1","student 2","student 3"))
df$student_id <- row.names(df)
library('reshape2')
df2 <- melt(df, id.vars = c('student_id','actual_exam_date'),
                variable.name = 'pretest_date',
                value.name = 'pretest_score')
df2 <- df2[!is.na(df2$pretest_score),]
df2$actual_exam_date <- as.Date(df2$actual_exam_date)
df2$pretest_date <- as.Date(df2$pretest_date)
df2$days_before_exam <- as.integer(df2$actual_exam_date - df2$pretest_date)
df2$pretest_score <- as.numeric(df2$pretest_score)
df2

The way I was able to calculate the maximum scores for each student was this:

aggregate(pretest_score ~ student_id, df2, max)

Now I am looking to identify the first and last pretest scores for each student, to calculate the difference between them. Is there a way to do it using aggregate?

Answer 1

Just another suggestion using data.table which allows you both order (if needed) and get the result without specifying anonymous functions and in a single line

library(data.table)
setDT(df2)[order(pretest_date), diff(pretest_score[c(1, .N)]), keyby = student_id]
#    student_id  V1
# 1:  student 1 100
# 2:  student 2 160
# 3:  student 3  90

Answer 2

First order your data by pretest date with

df2 <- df2[order(df2$pretest_date), ]

and then

aggregate(pretest_score ~ student_id, df2, function(x) tail(x, 1) - x[1])
  student_id pretest_score
1  student 1           100
2  student 2           160
3  student 3            90

Answer 3

First value for each student

> aggregate(pretest_score ~ student_id, df2, head, 1)
  student_id pretest_score
1  student 1           300
2  student 2           260
3  student 3           300

Last value for each student

  > aggregate(pretest_score ~ student_id, df2, tail, 1)
      student_id pretest_score
    1  student 1           400
    2  student 2           420
    3  student 3           390

Answer 4

There are several ways of doing what you are asking. With dplyr you can find the minimum and the maximum

 df2 %>% group_by(student_id) %>%
         filter(pretest_score == max(pretest_score) | pretest_score == min(pretest_score)) %>%
         mutate(differ = max(pretest_score) - min(pretest_score))
Source: local data frame [6 x 6]
Groups: student_id [3]

  student_id actual_exam_date pretest_date pretest_score days_before_exam differ
       (chr)           (date)       (date)         (dbl)            (int)  (dbl)
1  student 1       2013-04-30   2013-03-13           230               48    170
2  student 2       2013-05-30   2013-03-13           260               78    160
3  student 3       2013-08-30   2013-03-13           230              170    170
4  student 1       2013-04-30   2013-04-24           400                6    170
5  student 3       2013-08-30   2013-04-24           400              128    170
6  student 2       2013-05-30   2013-05-12           420               18    160

while with aggregate:

 aggregate(df2$pretest_score, list(df2$student_id), FUN = function(x) max(x) - min(x))
    Group.1   x
1 student 1 170
2 student 2 160
3 student 3 170

Answer 5

I think this one is easier with tapply:

tapply(df2$pretest_score, df2$student_id, function(x) tail(x,1)-head(x,1))

> student 1 student 2 student 3
        100       160        90