Hello, I have a data set like one below.
First,
I'd like to replace the empty cells with NA, and then the one immediately
above.
I could replace NAs with the immediate one, but don't know for the empty
cells.
index <- which(is.na(data1$year))
while (any(index)) {
dummy$data1[index] <- dummy$data1[index - 1]
index <- which(is.na(data1$year))
}
Second,
The id in fact represents year-month-date of sampling (i.e., 10104 means
2001-01-04).
How can I split the id column into three columns of year, month, and date,
respectively such that I can do time-series analysis after aggregating them
for each month?
Or, Is there a way to transform the id into Date for doing irregular ts
analysis?
With much thanks,
Keun-Hyung
> data1
year month id
1 2001 Jan 10104
2 NA 10110
3 NA 10116
4 NA 10122
5 NA Feb 10201
6 NA 10208
7 NA 10216
8 NA 10226
9 NA Mar 10303
10 NA 10309
50 2002 Jan 20104
51 NA 20111
52 NA 20117
53 NA 20124
54 NA Feb 20201
55 NA 20207
56 NA 20214
57 NA 20227
58 NA Mar 20306
59 NA 20313
60 NA 20320
94 2003 Jan 30103
95 NA 30113
96 NA 30122
97 NA 30128
98 NA Feb 30204
99 NA 30214
100 NA 30219
101 NA Mar 30307
102 NA 30313
103 NA 30317
104 NA 30324
> str(data1)
'data.frame': 32 obs. of 3 variables:
$ year : int 2001 NA NA NA NA NA NA NA NA NA ...
$ month: chr "Jan" "" "" "" ...
$ id : int 10104 10110 10116 10122 10201 10208 10216 10226 10303 10309
...
[[alternative HTML version deleted]]
