不规则时间序列中的for循环

Question

I'm looking for advice on how to loop through the following, which is a subset of a much larger data set. I hope the following representation works.

mydf <- structure(list(site_id = c("39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", "39ADA00070", 
"39ADA00070", "39ADA00070", "39ALL00184", "39ALL00184", "39ALL00184", 
"39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", 
"39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", 
"39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", 
"39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184", "39ALL00184"
), date = structure(c(6339, 8594, 9293, 9441, 10014, 10604, 11080, 
11821, 12717, 12907, 13081, 13277, 13459, 13635, 13822, 14012, 
14207, 14207, 14355, 14564, 14704, 14917, 15105, 15271, 15478, 
15644, 15833, 15834, 16009, 16203, 7783, 8406, 8554, 8686, 9034, 
9260, 9632, 9777, 10002, 10491, 10491, 11060, 11585, 12145, 12145, 
12696, 13242, 13242, 13775, 14363, 14881, 15428, 15974), class = "Date"), 
    var1 = c(NA, NA, NA, NA, NA, NA, NA, NA, NA, 159L, 148L, 
    149L, 134L, 179L, 205L, 193L, 109L, 109L, 177L, 75L, 272L, 
    150L, 115L, 232L, 230L, 183L, 159L, 159L, 304L, 220L, NA, 
    NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    -98L, -98L, -38L, -74L, -74L, -80L, -48L), var2 = c(NA, NA, 
    NA, NA, NA, NA, NA, NA, 16.8, 16.8, 14.5, 14.2, 15.1, 14.5, 
    15, 15.2, 13.2, 13.2, 15, 15.2, 15.1, 14.4, 14.8, 15.2, 16.3, 
    NA, 14.3, 14.3, 15.6, 14.8, NA, 12, 14.7, NA, 14.6, NA, 13.7, 
    12.3, 12.5, 13.5, 13.5, 12.5, 13.1, 14.2, 14.2, 14.1, 12.5, 
    12.5, 13.5, 12.7, 12.6, 12.5, 12.6), var3 = c(NA, NA, NA, 
    NA, NA, NA, NA, NA, 7.35, 7.85, 7.5, 7.47, 7.62, 7.08, 7.08, 
    7.2, 7.4, 7.4, 7.26, 7.05, 6.56, 7.2, 7.42, 6.5, 7.81, 8.43, 
    7.57, 7.57, 7.42, 7.72, NA, 6.58, 6.8, NA, 7.75, NA, 7.06, 
    6.77, 6.41, 6.84, 6.84, 7.85, 7.13, 7.26, 7.26, 7.06, 7.14, 
    7.14, 7.11, 6.9, 7.11, 7.2, 7.1), var4 = c(NA, 283L, 216L, 
    223L, 256L, 165L, 192L, 216L, 173L, 216L, 179L, 282L, 146L, 
    227L, 141L, 210L, 160L, 162L, 157L, 140L, 235L, 166L, 216L, 
    NA, 162L, 193L, 286L, 274L, 163L, 209L, NA, 304L, 321L, 293L, 
    398L, 302L, 301L, 282L, 288L, 292L, 292L, 302L, 515L, 309L, 
    309L, 323L, 338L, 295L, 280L, 279L, 325L, 328L, 322L), var5 = c(NA, 
    NA, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 
    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 
    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2), var6 = c(NA, NA, 
    29L, 32L, 36L, 24L, 25L, 29L, 27L, 27L, 24L, 32L, 21L, 27L, 
    21L, 26L, 23L, 24L, 25L, 20L, 24L, 22L, 28L, 24L, 20L, 23L, 
    30L, 29L, 21L, 24L, 15L, 15L, 18L, 15L, 15L, 15L, 15L, 15L, 
    15L, 15L, 15L, 15L, 15L, 15L, 15L, 15L, 15L, 15L, 15L, 15L, 
    15L, 15L, 15L), var7 = c(NA, NA, 77, 83, 87, 66, 73, 73, 
    65, 76, 69, 93, 60, 76, 56, 77, 67, 68, 68, 60, 67, 63, 82, 
    69, 56, 68, 85, 83, 59, 68.2, 157, 159, 164, 169, 155, 176, 
    156, 156, 162, 162, 162, 160, 180, 163, 163, 158, 168, 171, 
    162, 167, 177, 167, 168), var8 = c(NA, NA, 25, 26, 29, 21, 
    22, 23, 20, 23, 21, 30, 17, 24, 16, 23, 20, 20, 21, 17, 23, 
    18, 25, 20, 17, 21, 27, 27, 17, 20.9, 91, 89, 96, 92, 86, 
    100, 89, 91, 92, 94, 94, 91, 97, 91, 91, 92, 98, 99, 94, 
    100, 106, 98, 100), var9 = c(1.02, 1, 0.37, 0.48, 0.88, 0.16, 
    0.17, 0.24, 0.25, 5.98, 0.26, 0.54, 0, 0.19, 0, 0.18, 0.14, 
    0.13, 0.16, 0.11, 0.19, 0.16, 0.26, NA, 0.11, 0.27, 0.19, 
    0.19, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, NA, 0.1, 0.1, 0.1, 
    0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0, 0, 0.1, 0.1, 
    0.1), var10 = c(50, 48, 64, 55, 52, 64, 69, 63.3, 56.1, 40.6, 
    58.6, 43.9, 62.2, 51.9, 55.6, 53.4, 61.3, 61, 61.1, 61.9, 
    51.5, 60.7, 52.2, NA, 66, 52.8, 46.8, 47.5, 59.2, 53.4, NA, 
    560, 650, 540, 548, 655, 565, 531, 540, 501, 501, 531, 535, 
    547, 547, 492, 537, 542, 512, 542, 548, 581, 540)), class = "data.frame", row.names = c(NA, 
-53L), .Names = c("site_id", "date", "var1", "var2", "var3", 
"var4", "var5", "var6", "var7", "var8", "var9", "var10"))

This data.frame is a set of irregular time series using site_id as the main ID factor, date as the date, and then 10 variables. The actual data.frame has hundreds of IDs and dozens of factors.

I know I can access each time series by site_id using, for example

mydf[mydf$site_id == '39ADA00070', ][,3]

to get var1 for the first site_id .

What I am looking for is a robust for loop to run through the data.frame :

for (i in 1:length(site_id)){

  perform something on 
      var1 through var10

  output matrix of that something
}

That something would be any number of tests or plots, eg

GetOutliers() (from the extremevalues package) various plots, from ggplot2 etc, etc.

But first, I just need help getting the assignment of the for loop indices correct.

I am not against using apply ( ddply ) tools for this work, but I wanted to start with a basic for loop. Then I can clean up by addressing NA s, censored values, etc.

Thanks so much!

Answer 1

Try following:

for(ss in unique(mydf$site_id)){
    for(cc in 3:12){
        # do whatever function
        print(max(mydf[mydf$site_id == ss, cc],na.rm=TRUE))
    }
}

[1] 304
[1] 16.8
[1] 8.43
[1] 286
[1] 2
[1] 36
[1] 93
[1] 30
[1] 5.98
[1] 69
[1] -38
[1] 14.7
[1] 7.85
[1] 515
[1] 2
[1] 18
[1] 180
[1] 106
[1] 0.1
[1] 655