Home/ Questions/Q 8985295
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-15T21:17:57+00:00

Thanks in advance for your help. I am using R and let’s say that

  • 0

Thanks in advance for your help.
I am using R and let’s say that I have a datatable (or eventually a timeseries with zoo) with the following format:

Col1: time   Values
Day1 H1      Value
Day1 H2      Value
Day1 H3      Value
Day1 H4      Value

Day2 H1      Value
Day2 H2      Value
Day2 H3      Value
Day2 H4      Value

Day3 H1      Value
...

Let’s say that I would like to construct a matrix with the following format:
Rows:Days

    H1       H2       H3       H4
D1  Values   Values   Values   Values
D2  Values   Values   Values   Values
D3  Values   Values   Values   Values

and also:

    average(H1,H2)       average(H3,H4)
D1  Values               Vales
D2  Values               Vales
D3  Values               Vales

In some languages such as C++ we would probably proceed with a double ‘for’ but I am not sure it is the best way to proceed here.
Thanks a lot, I am new to R and I am quite lost with the different logic (but very interesting one).

I have seen other questions on the topic but I am not clear at all.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    This can be done with some basic reshape work, and aggregate() or within() for the means:

    First, some sample data is very helpful:

    set.seed(1)
temp <- data.frame(Col1 = paste("Day", rep(1:4, each = 4), sep=""),
                   times = paste("H", rep(1:4, times = 4), sep=""),
                   Values = runif(16, min=0, max=10))
temp
#    Col1 times    Values
# 1  Day1    H1 2.6550866
# 2  Day1    H2 3.7212390
# 3  Day1    H3 5.7285336
# 4  Day1    H4 9.0820779
# 5  Day2    H1 2.0168193
# 6  Day2    H2 8.9838968
# 7  Day2    H3 9.4467527
# 8  Day2    H4 6.6079779
# 9  Day3    H1 6.2911404
# 10 Day3    H2 0.6178627
# 11 Day3    H3 2.0597457
# 12 Day3    H4 1.7655675
# 13 Day4    H1 6.8702285
# 14 Day4    H2 3.8410372
# 15 Day4    H3 7.6984142
# 16 Day4    H4 4.9769924

    Second, use reshape to go from long to wide format

    tempwide <- reshape(temp, direction = "wide", idvar="Col1", timevar="times")
tempwide
#    Col1 Values.H1 Values.H2 Values.H3 Values.H4
# 1  Day1  2.655087 3.7212390  5.728534  9.082078
# 5  Day2  2.016819 8.9838968  9.446753  6.607978
# 9  Day3  6.291140 0.6178627  2.059746  1.765568
# 13 Day4  6.870228 3.8410372  7.698414  4.976992

    Third, use rowMeans on the desired subset of your columns. You can also use aggregate if you prefer, but this is a convenient way to transform your original data.frame.

    tempwide <- within(tempwide, {
  mean.H1H2 <- rowMeans(tempwide[2:3])
  mean.H3H4 <- rowMeans(tempwide[4:5])
})
tempwide
#    Col1 Values.H1 Values.H2 Values.H3 Values.H4 mean.H3H4 mean.H1H2
# 1  Day1  2.655087 3.7212390  5.728534  9.082078  7.405306  3.188163
# 5  Day2  2.016819 8.9838968  9.446753  6.607978  8.027365  5.500358
# 9  Day3  6.291140 0.6178627  2.059746  1.765568  1.912657  3.454502
# 13 Day4  6.870228 3.8410372  7.698414  4.976992  6.337703  5.355633
    • 0