Let's say we have sequences that consist of 5 different events/states (A-E) like this:
library(TraMineR)
data(actcal)
actcal.seq <- seqdef(actcal, 13:24, alphabet=c("A","B","C","D","E")
Is it possible to now create a subset of actcal.seq that only contains for instance events A, C and E? If yes, then how is this done?
Clarification: I want to extract any sequence that contains A, C or E. If any of those contain B or D those events should be removed from the returned sequence. For instance, a sequence A-A-B-C-C-D-E-E should be returned as A-A-C-C-E-E.
Clarification 2: The input sequences should use the alphabet=c("A","B","C","D","E") while the modified sequence object I'm looking for should use the alphabet=c("A","C","E"). Some more examples as requested are given below:
"A-B-C-D-E" => "A-C-E"
"A-C-A-E" => "A-C-A-E"
"B-D" => NA or ""
"B-D-B-A-D" => "A"
I'll appreciate any solution on how to solve this without having to re-read a subset of the data from database.
You can recode states B and D as missing by means of the seqrecode function. The default symbol used for missing is *. I illustrate using only the first 10 sequences of actcal
data(actcal)
actcal.seq <- seqdef(actcal[1:10,13:24], alphabet=c("A","B","C","D","E"))
## Recode B and D as *, the default missing symbol
actcal.rec.seq <- seqrecode(actcal.seq,
recodes = list("*"=c("B","D")), otherwise=NULL)
actcal.seq
# Sequence
# 2848 B-B-B-B-B-B-B-B-B-B-B-B
# 1230 D-D-D-D-A-A-A-A-A-A-A-D
# 2468 B-B-B-B-B-B-B-B-B-B-B-B
# 654 C-C-C-C-C-C-C-C-C-B-B-B
# 6946 A-A-A-A-A-A-A-A-A-A-A-A
# 1872 D-B-B-B-B-B-B-B-B-B-B-B
# 2905 D-D-D-D-D-D-D-D-D-D-D-D
# 106 A-A-A-A-A-A-A-A-A-A-A-A
# 5113 A-A-A-A-A-A-A-A-A-A-A-A
# 4503 A-A-A-A-A-A-A-A-A-A-A-A
actcal.rec.seq
# Sequence
# 2848 *-*-*-*-*-*-*-*-*-*-*-*
# 1230 *-*-*-*-A-A-A-A-A-A-A-*
# 2468 *-*-*-*-*-*-*-*-*-*-*-*
# 654 C-C-C-C-C-C-C-C-C-*-*-*
# 6946 A-A-A-A-A-A-A-A-A-A-A-A
# 1872 *-*-*-*-*-*-*-*-*-*-*-*
# 2905 *-*-*-*-*-*-*-*-*-*-*-*
# 106 A-A-A-A-A-A-A-A-A-A-A-A
# 5113 A-A-A-A-A-A-A-A-A-A-A-A
# 4503 A-A-A-A-A-A-A-A-A-A-A-A
Dropping the missing states
actcal.rec.comp.seq <- seqdef(actcal.rec.seq,
left="DEL", gap="DEL", right="DEL",
missing="*", alphabet=c("A","C","E"))
Removing sequences that contain only missing
(rec.seq <- actcal.rec.comp.seq[!is.na(seqdur(actcal.rec.comp.seq)[,1]),])
# Sequence
# 2103 A-A-A-A-A-A-A-A-A-A-A-A
# 3972 C-C-C-C-C-C-C-C-C
# 5238 C
# 4977 C-C-C-C-C-C-C-C-C-C-C-C
# 528 A-A-A-A-A-A-A-A-A-A-A-A
And if you want only the sequence of distinct successive states
seqdss(rec.seq)
# Sequence
# 2103 A
# 3972 C
# 5238 C
# 4977 C
# 528 A