SPSSX Discussion

create variables

Classic

List

Threaded

4 messages Options

abdelrhman elmubarak

create variables

I have data of 3 variables ID , center and date of visit … ID is unique

I want to create more 3 variables only for those who visit center A, as below

id	center	date of visit	repeat visit for center A	visit other center	lost customer
A	A	20090109	1st visit
A	A	20090319	yes	no	no
B	A	20090212	1st visit
B	C	20090401	no	yes	no
C	A	20090625	1st visit
C	C	20090731	no	no	yes
D	B	20090320
E	D	20090411
F	D	20090516
G	B	20090605

Thanks in advance

Abdulrahman Yousef Elmubark

Windows Live: Keep your friends up to date with what you do online.

Ruben Geert van den Berg

Re: create variables

Dear Abdulrahman,

I did not quite understand what you mean by 'lost customer'. Why isn't person B a lost customer (for center A, that is)? I think the SPSS syntax below will create the other two variables. I hope this is at least part of what you were looking for.

Kind regards,

Ruben van den Berg

*Read in the data.

Data list free/id_2 center_2 visit_2 (3a8).
Begin data
A A 20090109
A A 20090319
B A 20090212
B C 20090401
C A 20090625
C C 20090731
D B 20090320
E D 20090411
F D 20090516
G B 20090605
End data.

*Make numeric variables of strings.

AUTORECODE
VARIABLES=id_2 center_2 /INTO id center.

*Make data variable of date.

comp #yyyy=num(sub(visit_2,1,4),f4).
comp #mm=num(sub(visit_2,5,2),f2).
comp #dd=num(sub(visit_2,7,2),f2).
compute visit= DATE.MDY(#mm,#dd,#yyyy).
for visit(date).

*Aggregate within person and center.

sor cas by id center.
agg
/out * mod add
/bre id center
/fre=n
/lastv=max(visit)
/firstv=min(visit).
exe.

*Aggregate within person.

agg
/out * mod add
/bre id
/vis_oth=sd(center).
exe.

*Create repeated visit variable.

comp rep_vis=0.
if center=1 and visit=firstv rep_vis=1.
if center ne 1 rep_vis=2.
if center=1 and visit ne firstv rep_vis=3.
exe.

*Label repeated visit variable.

var lab rep_vis 'Repeated visit to center A?'.
val lab rep_vis 1'First visit to center a' 2'No' 3'Repeated visit to center a'.

*Create 'visit other center' variable.

comp vis_other=0.
if vis_oth gt 0 vis_other=1.
exe.

*Label 'visit other center' variable.

var lab vis_other 'Did the respondent visit more than one center?'.
val lab vis_other 0 'Has visited only one center' 1'Has visited two or more centers'.

*Delete all variables except the relevant ones.

matc fil fil *
/kee id center visit rep_vis vis_other.
exe.

Date: Tue, 3 Nov 2009 08:23:54 +0000
From: [hidden email]
Subject: create variables
To: [hidden email]

I have data of 3 variables ID , center and date of visit … ID is unique

I want to create more 3 variables only for those who visit center A, as below

id	center	date of visit	repeat visit for center A	visit other center	lost customer
A	A	20090109	1st visit
A	A	20090319	yes	no	no
B	A	20090212	1st visit
B	C	20090401	no	yes	no
C	A	20090625	1st visit
C	C	20090731	no	no	yes
D	B	20090320
E	D	20090411
F	D	20090516
G	B	20090605

Thanks in advance

Abdulrahman Yousef Elmubark

Windows Live: Keep your friends up to date with what you do online.

Express yourself instantly with MSN Messenger! MSN Messenger

abdelrhman elmubarak

Re: create variables

Dear Ruben

Thank you very much for your syntax …this exactly what I am looking for first tow variables.

For lost customer, Sorry it is my problem I didn’t explain well . I mean if customer didn’t shows up again in any center after his first visit .

What will happen if I removed customer C visit to center C? Will this help in explaining lost customer.

Thanks again

Abdulrahman

Date: Tue, 3 Nov 2009 12:59:24 +0000
From: [hidden email]
Subject: Re: create variables
To: [hidden email]

Dear Abdulrahman,

I did not quite understand what you mean by 'lost customer'. Why isn't person B a lost customer (for center A, that is)? I think the SPSS syntax below will create the other two variables. I hope this is at least part of what you were looking for.

Kind regards,

Ruben van den Berg

*Read in the data.

Data list free/id_2 center_2 visit_2 (3a8).
Begin data
A A 20090109
A A 20090319
B A 20090212
B C 20090401
C A 20090625
C C 20090731
D B 20090320
E D 20090411
F D 20090516
G B 20090605
End data.

*Make numeric variables of strings.

AUTORECODEVARIABLES=id_2 center_2 /INTO id center.

*Make data variable of date.

comp #yyyy=num(sub(visit_2,1,4),f4).
comp #mm=num(sub(visit_2,5,2),f2).
comp #dd=num(sub(visit_2,7,2),f2).
compute visit= DATE.MDY(#mm,#dd,#yyyy).
for visit(date).

*Aggregate within person and center.

sor cas by id center.
agg
/out * mod add
/bre id center
/fre=n
/lastv=max(visit)
/firstv=min(visit).
exe.

*Aggregate within person.

agg
/out * mod add
/bre id
/vis_oth=sd(center).
exe.

*Create repeated visit variable.

comp rep_vis=0.
if center=1 and visit=firstv rep_vis=1.
if center ne 1 rep_vis=2.
if center=1 and visit ne firstv rep_vis=3.
exe.

*Label repeated visit variable.

var lab rep_vis 'Repeated visit to center A?'.
val lab rep_vis 1'First visit to center a' 2'No' 3'Repeated visit to center a'.

*Crea! te 'visit other center' variable.

comp vis_other=0.
if vis_oth gt 0 vis_other=1.
exe.

*Label 'visit other center' variable.

var lab vis_other 'Did the respondent visit more than one center?'.
val lab vis_other 0 'Has visited only one center' 1'Has visited two or more centers'.

*Delete all variables except the relevant ones.

matc fil fil *
/kee id center visit rep_vis vis_other.
exe.

Date: Tue, 3 Nov 2009 08:23:54 +0000
From: [hidden email]
Subject: create variables
To: [hidden email]

I have data of 3 variables ID , center and date of visit … ID is unique

I want to create more 3 variables only for those who visit center A, as below

id	center	date of visit	repeat visit for center A	visit other center	lost customer
A	A	20090109	1st visit
A	A	20090319	yes	no	no
B	A	20090212	1st visit
B	C	20090401	no	yes	no
C	A	20090625	1st visit
C	C	20090731	no	no	yes
D	B	20090320
E	D	20090411
F	D	20090516
G	B	20090605

Thanks in advance

Abdulrahman Yousef Elmubark

Windows Live: Keep your friends up to date with what you do online.

Express yourself instantly with MSN Messenger! MSN Messenger

Windows Live: Friends get your Flickr, Yelp, and Digg updates when they e-mail you.

Ruben Geert van den Berg

Re: create variables

Dear Abdulrahman,

I'm sorry but I still don't really understand what you mean. Perhaps you could send more sample data (more cases, more patterns, less 'system missings') to further clarify?

Kind regards,

Ruben van den Berg

Date: Tue, 3 Nov 2009 13:46:44 +0000
From: [hidden email]
Subject: Re: create variables
To: [hidden email]

Dear Ruben

Thank you very much for your syntax …this exactly what I am looking for first tow variables.

For lost customer, Sorry it is my problem I didn’t explain well . I mean if customer didn’t shows up again in any center after his first visit .

What will happen if I removed customer C visit to center C? Will this help in explaining lost customer.

Thanks again

Abdulrahman

Date: Tue, 3 Nov 2009 08:23:54 +0000
From: [hidden email]
Subject: create variables
To: [hidden email]

I have data of 3 variables ID , center and date of visit … ID is unique

I want to create more 3 variables only for those who visit center A, as below

id	center	date of visit	repeat visit for center A	visit other center	lost customer
A	A	20090109	1st visit
A	A	20090319	yes	no	no
B	A	20090212	1st visit
B	C	20090401	no	yes	no
C	A	20090625	1st visit
C	C	20090731	no	no	yes
D	B	20090320
E	D	20090411
F	D	20090516
G	B	20090605

Thanks in advance

Abdulrahman Yousef Elmubark

Windows Live: Keep your friends up to date with what you do online.

Express yourself instantly with MSN Messenger! MSN Messenger

Windows Live: Friends get your Flickr, Yelp, and Digg updates when they e-mail you.

Express yourself instantly with MSN Messenger! MSN Messenger