(no subject)

[Charley Trimble]

Help, please.
 
I'm wanting to create new variables for each defined geographical area (a combination zip code_city description thru a variable named cz_city below) that ties individual sales to the zip code_city location of each customer and which also tags the corresponding employee (emp) handling the sale.  The listing below, for example, creates three new variables (z12345_Dallas, z77239_Miami, and z24596_Pittsb), which reveal the location of the sale by zip code_city, the amount of dollarsales (sales$), and the employee responsible for garnering the sales.  The syntax for creating these three variables is listed below, as well.
 
With that said, there may be, in reality, as many as 500-600 zip codes by state and, what is more, the number of individual sales by employee may be in the millions.
 
How can I create new variables without going through the graphical user interface of SPSS for each and every COMPUTE and IF command for each zip code_city variable?  Note the zip codes in the illustration below are imaginary but I do, nonetheless, have all of the zip codes in the U.S. and have matched them with the corresponding city name, e.g. z77239_Miami to create the cz_city variable.  The question is how can I more quickly create these variables?
 
Can someone please educate me on the syntax required to create these new zvariables, as well as the corresonding sysmis commands, which are also singularly noted below?  Thank you very, very much.
Charley Trimble
 
 
emp     id     c_zip     cz_city             sales$     z12345_Dallas     z77233_Miami     z24596_Pittsb
Bob     100    12345    z12345_Dallas         4423.13    4423.13                    .00            .00
Bob     100    12345    z12345_Dallas         3952.12    3952.12                    .00            .00
Bob     200    77233    z77233_Miami          2100.19        .00                2100.19            .00
Bob     300    24596    z24596_Pittsb         2129.74        .00                    .00        2129.74
Sally   100    24596    z24596_Pittsb         2998.24        .00                    .00        2998.24
Sally   300    12345    z12345_Dallas         3219.80    3219.80                    .00            .00
Susan   200    77233    z77233_Miami          4290.04        .00                4290.04            .00
Susan   200    12345    12345_Dallas          5679.49    5679.49                    .00            .00
 
Number of cases read: 8 Number of cases listed: 8
 
 
 
IF  (c_zip = 12345) z12345_Dallas=sales$.
VARIABLE LABELS  z12345_Dallas 'z12345 Dallas'.

IF  (c_zip = 77233) z77233_Miami=sales$.
VARIABLE LABELS  z77233_Miami 'Miami Sales'.

IF  (c_zip = 24596) z24596_Pittsb=sales$.
VARIABLE LABELS  z24596_Pittsb 'Pittsburgh sales'.
EXECUTE.
 
if (sysmis (z12345_Dallas)) z12345_Dallas = 0.
if (sysmis (z77233_Miami)) z77233_Miami = 0.
if (sysmis (z24596_Pittsb)) z24596_Pittsb = 0.
EXECUTE .

====================To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Hector Maletta]

I think, Charley, you are approaching this the wrong way. I imagine your
original data set has only the first five variables (emp, id, c_zip,
cz_city, and sales$. This master file could be handled in many ways to
generate output tables and/or partial files concerning sales by particular
employees, or sales in a particular city, or whatever combination you may
fancy. You don't actually need to create different variables for sales in
Dallas, sales in Miami and so on.
So it all depends on what you actually need. If you need output tables
showing total sales by salesmen in Dallas, you may generate them quite
easily. You may also need total sales by city, or whatever. For things like
these, you do not need new variables: you just order the tables you need.
For instance:
MEANS SALES$ BY CZ_CITY BY EMP /CELLS SUM.
This generates a table showing total sales for each employee in each city.
Besides the sum, you may ask for the mean, the maximum and minimum sale, the
standard deviation, and several other measures regarding the sales of each
employee in each city. You may also omit the employee by stating only MEANS
SALES$ BY CZ_CITY/CELLS SUM. If you omit the CELLS SUM subcommand, the
default statistics for each cell are the mean, the std deviation and the
number of cases.
Other commands are available, such as custom tables, which require specific
modules outside the Base system, and give you more possibilities.

Hector




-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Charley Trimble
Sent: 05 July 2008 20:18
To: [hidden email]
Subject:

Help, please.

I'm wanting to create new variables for each defined geographical area (a
combination zip code_city description thru a variable named cz_city below)
that ties individual sales to the zip code_city location of each customer
and which also tags the corresponding employee (emp) handling the sale.  The
listing below, for example, creates three new variables (z12345_Dallas,
z77239_Miami, and z24596_Pittsb), which reveal the location of the sale by
zip code_city, the amount of dollarsales (sales$), and the employee
responsible for garnering the sales.  The syntax for creating these three
variables is listed below, as well.

With that said, there may be, in reality, as many as 500-600 zip codes by
state and, what is more, the number of individual sales by employee may be
in the millions.

How can I create new variables without going through the graphical user
interface of SPSS for each and every COMPUTE and IF command for each zip
code_city variable?  Note the zip codes in the illustration below
are imaginary but I do, nonetheless, have all of the zip codes in the U.S.
and have matched them with the corresponding city name, e.g. z77239_Miami to
create the cz_city variable.  The question is how can I more quickly create
these variables?

Can someone please educate me on the syntax required to create these new
zvariables, as well as the corresonding sysmis commands, which are also
singularly noted below?  Thank you very, very much.
Charley Trimble


emp     id     c_zip     cz_city             sales$     z12345_Dallas
z77233_Miami     z24596_Pittsb
Bob     100    12345    z12345_Dallas         4423.13    4423.13
        .00            .00
Bob     100    12345    z12345_Dallas         3952.12    3952.12
        .00            .00
Bob     200    77233    z77233_Miami          2100.19        .00
    2100.19            .00
Bob     300    24596    z24596_Pittsb         2129.74        .00
        .00        2129.74
Sally   100    24596    z24596_Pittsb         2998.24        .00
        .00        2998.24
Sally   300    12345    z12345_Dallas         3219.80    3219.80
        .00            .00
Susan   200    77233    z77233_Miami          4290.04        .00
    4290.04            .00
Susan   200    12345    12345_Dallas          5679.49
5679.49                    .00            .00

Number of cases read: 8 Number of cases listed: 8



IF  (c_zip = 12345) z12345_Dallas=sales$.
VARIABLE LABELS  z12345_Dallas 'z12345 Dallas'.

IF  (c_zip = 77233) z77233_Miami=sales$.
VARIABLE LABELS  z77233_Miami 'Miami Sales'.

IF  (c_zip = 24596) z24596_Pittsb=sales$.
VARIABLE LABELS  z24596_Pittsb 'Pittsburgh sales'.
EXECUTE.

if (sysmis (z12345_Dallas)) z12345_Dallas = 0.
if (sysmis (z77233_Miami)) z77233_Miami = 0.
if (sysmis (z24596_Pittsb)) z24596_Pittsb = 0.
EXECUTE .


To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD