[Traminer-users] within dyad distances

[Hadrien Commenges]

I don't know if there is an easier way, already implemented in the Traminer package, but a base-R solution would be: 

1. splitting your data - split() - and get the results in a list, each slot of the list storing one individual's sequences 
2. applying a function to each slot of the list with lapply() to compute the distances for each individual 

If you do that, it is also very easy to compute with multiple cores using mclapply() instead of lapply(). 

Regards, 

Hadrien 


De: "Reynolds, Jeremy E" <reyno113 at purdue.edu> 
À: "traminer-users" <traminer-users at lists.r-forge.r-project.org> 
Envoyé: Vendredi 7 Février 2020 19:05:01 
Objet: [Traminer-users] within dyad distances 



Dear Traminer Users, 



I would like to compute distances between sequences that belong to the same person for every person in my data. 

The code below seems to work: it calculates the distances between the expected and actual work schedule for each of the 10 people in the data. 



My code, however, is terribly inefficient. For instance, it calculates the entire pairwise distance matrix and then overwrites most of it with NA. 

This leaves me with two questions: 



1) Does Traminer have a way to calculate just the distances between sequences that below to the same person? 

(e.g., with clever use of the refseq option in the seqdist command or with the seqdistmc command for multi-channel sequence analysis) 



2) Is there a way to extract the elements just below the diagonal without overwriting all the other values with NA? 



Thanks, 



Jeremy 





mymat <- rbind( 

c(1,1,0,0,1,1,1,1,1,0,0,0),c(1,2,0,0,1,1,1,1,1,1,0,0), 

c(2,1,0,0,1,1,1,1,1,0,0,0),c(2,2,0,0,1,1,1,1,1,1,1,0), 

c(3,1,0,0,1,1,1,1,1,0,0,0),c(3,2,0,0,1,1,1,1,1,1,1,1), 

c(4,1,0,0,1,1,1,1,1,0,0,0),c(4,2,0,1,1,1,1,1,1,0,0,0), 

c(5,1,0,0,1,1,1,1,1,0,0,0),c(5,2,1,1,1,1,1,1,1,0,0,0), 

c(6,1,0,0,1,1,1,1,1,0,0,0),c(6,2,0,0,0,1,1,1,1,0,0,0), 

c(7,1,0,0,1,1,1,1,1,0,0,0),c(7,2,0,0,0,0,1,1,1,0,0,0), 

c(8,1,0,0,1,1,1,1,1,0,0,0),c(8,2,0,0,0,0,0,1,1,0,0,0), 

c(9,1,0,0,1,1,1,1,1,0,0,0),c(9,2,0,0,0,0,0,0,1,0,0,0), 

c(10,1,0,0,1,1,1,1,1,0,0,0),c(10,2,0,0,0,0,0,0,0,0,0,0) 

) 



colnames(mymat) <- c("ID", "sched", "t1", "t2", "t3", "t4", "t5", "t6", "t7", "t8", "t9", "t10") 

mymat <- as.data.frame(mymat) 

mymat$sched <- factor(mymat$sched,levels = c(1,2), labels = c("Expected", "Actual")) 



library(TraMineR) 



# make sequence object 

labels <- c("working", "not working") 

scode <- c("W", "N") 

seq <- seqdef(mymat, 3:12, states = scode, labels = labels) 



# sequence index plot 

seqIplot(seq, with.legend = T, main = "Expected and Actual Work Schedules of 5 People",border = NA) 



# calculate dynamic hamming distances for every possible pair 

distmat <- seqdist(seq, method = "DHD", indel = 1, sm = NULL) 



# extract the elements just below the main diagonal 

low <- 1 

high <- 1 

delta <- row(distmat) - col(distmat) 

distmat[delta < low | delta > high] <- NA 

distvec <- na.omit(as.data.frame(distmat[delta >= low | delta <= high])) 

#repeat the last entry 

distvec <- rbind(distvec,tail(distvec, n=1)) 



#attach the off diagonal to the data frame 

colnames(distvec) <- "dist" 

mymat <- cbind(mymat,distvec) 







Dr. Jeremy Reynolds 

Professor 

307 Stone Hall 

Department of Sociology 

700 W. State Street 

Purdue University 

West Lafayette, IN 47907 

Phone: (765) 496-3348 

[ https://cla.purdue.edu/directory/profiles/jeremy-reynolds.html | https://cla.purdue.edu/directory/profiles/jeremy-reynolds.html ] 


Pronouns: he/him/his 



_______________________________________________ 
Traminer-users mailing list 
Traminer-users at lists.r-forge.r-project.org 
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/traminer-users 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/traminer-users/attachments/20200209/0396ac14/attachment.html>