[Traminer-users] création d'une séquence à partir d'un format SPELL

Guillaume Ollivier gollivier at avignon.inra.fr
Wed Apr 4 11:42:08 CEST 2012


I’ve specified the status variable but the problem is still maintained.

I’ve also tried to use seqformat instead of seqdef with no results.

The solution comes from the transformation of status in factor to numeric
(status_num) that I put directly in the seqdef :

status_num <-  as.integer(cv$status)

cv <- data.frame(cv, status_num)

cv.seq <- seqdef(cv, id = "idcv", begin = "start", end = "end",
status='status_num', informat = 'SPELL', fillblanks="IN", overwrite=TRUE,
process=FALSE)

 

Moreover, another point explained the failure : I used Process=TRUE without
the pdata and pvar parameters.

 

Thanks a lot !

 

 

De : traminer-users-bounces at r-forge.wu-wien.ac.at
[mailto:traminer-users-bounces at r-forge.wu-wien.ac.at] De la part de Chris
Cameron
Envoyé : mardi 3 avril 2012 23:37
À : Users questions
Objet : Re: [Traminer-users] création d'une séquence à partir d'un format
SPELL

 

 

I think you need to also specify the status variable:

cv.seq <- seqdef(cv, var=c("idcv", "start", "end", "status"),
informat="SPELL", id = "idcv", begin = "start", end = "end", status='status'
)

 

If that does not work, pages 42-43 of the TraMineR 1.4 Users manual explains
how to deal with a status in factor format and convert to STS as suggested
by Hadrian. 

 

 

On Apr 3, 2012, at 1:00 PM, Guillaume Ollivier wrote:





Bonjour,

 

Je cherche à créer un seq à partir de données de CV formatées en SPELL.

str(cv) :

'data.frame':   444 obs. of  12 variables:

$ idcv          : num  8.04e+13 8.04e+13 8.04e+13 8.04e+13 8.04e+13 ...

$ start         : int  1989 1993 1994 1994 1995 1996 1997 1997 1998 1999 ...

$ end           : int  1994 1994 1994 1997 1997 1996 1999 1999 2004 2009 ...

$ status        : Factor w/ 7 levels "","academy","company",..: 1 1 1 1 7 7
1 1 2 1 ...

$ nome_completo : Factor w/ 49 levels "Adriano Arriel Saquet",..: 30 30 30
30 30 30 30 30 30 30 ...

$ ano_nascimento: int  1972 1972 1972 1972 1972 1972 1972 1972 1972 1972 ...

$ sexo          : Factor w/ 2 levels "Feminino","Masculino": 2 2 2 2 2 2 2 2
2 2 ...

$ etat          : Factor w/ 17 levels "","AC","AM","BA",..: 16 16 16 16 16
16 16 16 16 16 ...

$ pais          : Factor w/ 2 levels "","Brasil": 2 2 2 2 2 2 2 2 2 2 ...

$ duree_parcours: num  23 23 23 23 23 23 23 23 23 23 ...

$ ano_mini      : num  1989 1989 1989 1989 1989 ...

$ denschaves    : num  0.564 0.564 0.564 0.564 0.564 ...

----------------

 

Pour cela j’utilise la fonction seqdef où je spécifie la localisation de la
variable status dans lequel il y a des données manquantes:

cv.seq <- seqdef(cv, var=c("idcv", "start", "end", "status"),
informat="SPELL", id = "idcv", begin = "start", end = "end")

 

 

mais R me renvoie le message suivant :

[>] SPELL data converted into 49 STS sequences

[>] found missing values ('NA') in sequence data

[>] preparing 49 sequences

[>] coding void elements with '%' and missing values with '*'

[!] sequence with index:
1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,2
9,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49 contains only
missing values.

     This may produce inconsistent results.

[>] 0 distinct states appear in the data:

     1 = NA

     0 =

 [>] alphabet (state labels):

     1 = NA (NA)

     0 =  ()

[>] 49 sequences in the data set

[>] min/max sequence length: 100/100

-------------

 

Il semble que l’opération de conversion bloque  ce que confirme l’examen de
l’objet cv.seq  qui ne contient que des données manquantes. R ne reconnait
visiblement pas la variable status et les différents états qu’elle est
censée contenir.

Pourriez-vous me donner des pistes pour solutionner ce problème

 

Bien cordialement

 

Guillaume Ollivier

INRA

Plateformes AMAnDeS.Txt

"Appui Méthodologique en Analyse des Données Sociologiques et Textuelles"

UR 0767 Ecodéveloppement

Agroparc, 84914 Avignon cedex 9

tel : 0432722579

http://www4.paca.inra.fr/ecodeveloppement/Plateforme-Amandes.txt

 

_______________________________________________
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/20120404/6e022dc6/attachment.html>


More information about the Traminer-users mailing list