simulate data

[Moon Kid]

Is there a way or mechanism with SPSS to create "data" based on
existing variables and givin plausible ranges of their values.

I don't mean random numbers.

I want to simulate my analyse with some data.

For example two variables.
v1 is numeric with a range from 0 to 15 and a missing value of '-9'.
v2 is string with a max length of 64 and min length of 4 and missing
value ''.

Now I just need 100 observations (is it the correct SPSS word for
that?) with this givin parameters.

But the important point to know is that the variables itself (with all
its specifications, labes, etc) are still existing. I just want to fill
them. I don't want to create new variables while simulating data.

I created the variables while a GET DATA from a cvs-file and some
modifications (VALUE LABEL, etc) after that.

Is this possible?

=====================
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

[Moon Kid]

On 2014-03-26 21:23 Moon Kid <[hidden email]> wrote:
> Is this possible?

btw: Of course I found SIMPLAN but didn't unterstand it. I think
simulating data is quite simple task. variable, min, max, thats all.
SIMPLAN looks quite to complex for that.

=====================
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

[David Marso]

See LOOP and XSAVE commands!

Moon Kid wrote

Is there a way or mechanism with SPSS to create "data" based on
existing variables and givin plausible ranges of their values.

I don't mean random numbers.

I want to simulate my analyse with some data.

For example two variables.
v1 is numeric with a range from 0 to 15 and a missing value of '-9'.
v2 is string with a max length of 64 and min length of 4 and missing
value ''.

Now I just need 100 observations (is it the correct SPSS word for
that?) with this givin parameters.

But the important point to know is that the variables itself (with all
its specifications, labes, etc) are still existing. I just want to fill
them. I don't want to create new variables while simulating data.

I created the variables while a GET DATA from a cvs-file and some
modifications (VALUE LABEL, etc) after that.

Is this possible?

=====================
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

[Rick Oliver-3]

Well, if all you need to define is min and max:

input program.
loop #i=1 to 100.
compute var1=rv.uniform(min, max).
compute var2=rv.uniform(min, max).
etc.
end case.
end loop.
end file.
end input program.


If you want integer ranges: trunc(rv.uniform(min, max+1)).
Then use APPLY DICTIONARY to apply the variable definitions you indicated you already saved in another SAV file.
The variable names you specify in the input program should correspond to the variable names in the previously saved file.

I think you indicated previously that you did not want random values, but if the only parameters you are providing are min and max, then what else could the values between those parameters be?

Rick Oliver
Senior Information Developer
IBM Business Analytics (SPSS)
E-mail: [hidden email]



From:        Moon Kid <[hidden email]>
To:        [hidden email],
Date:        03/26/2014 03:48 PM
Subject:        Re: simulate data
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

On 2014-03-26 21:23 Moon Kid <[hidden email]> wrote:
> Is this possible?

btw: Of course I found SIMPLAN but didn't unterstand it. I think
simulating data is quite simple task. variable, min, max, thats all.
SIMPLAN looks quite to complex for that.

=====================
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

[Jon K Peck]

In order to create new rows in an existing dataset, you need to start with at least one case (observation, row).  Then transformation commands can generate a file with as many as needed.
For example (just one variable).  Typically you would fill these with, yes, random data.
data list list/x(f8.2).
begin data
1
end data.
loop #i = 1 to 10.
compute x = rv.uniform(0,15).
xsave outfile="c:\temp\tencases.sav".
end loop.
exec.

INPUT PROGRAMs can create entire new datasets with case values, but they can't do that into an existing dataset.  However, you could use APPLY DICTIONARY to copy over the metadata from your empty one.

There is a custom dialog called SimulateActiveDataset available from the SPSS Community website that can simulate a new dataset that is like an existing one, including the metadata, but it needs cases as input in order to determine the distributions to match up with.

Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: [phone]




From:        Moon Kid <[hidden email]>
To:        [hidden email],
Date:        03/26/2014 02:24 PM
Subject:        [SPSSX-L] simulate data
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

Is there a way or mechanism with SPSS to create "data" based on
existing variables and givin plausible ranges of their values.

I don't mean random numbers.

I want to simulate my analyse with some data.

For example two variables.
v1 is numeric with a range from 0 to 15 and a missing value of '-9'.
v2 is string with a max length of 64 and min length of 4 and missing
value ''.

Now I just need 100 observations (is it the correct SPSS word for
that?) with this givin parameters.

But the important point to know is that the variables itself (with all
its specifications, labes, etc) are still existing. I just want to fill
them. I don't want to create new variables while simulating data.

I created the variables while a GET DATA from a cvs-file and some
modifications (VALUE LABEL, etc) after that.

Is this possible?

=====================
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

[bdates]

Simulating data may take more than a minimum and maximum.  There are specifications for various distributions in the Transform Variable area.  Each requires different parameters.  You'll need to identify the distribution you want your data to follow.  In the last three months, there was a thread regarding generating datasets.  You can check it out in the archives.



-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Moon Kid
Sent: Wednesday, March 26, 2014 4:44 PM
To: [hidden email]
Subject: Re: simulate data

On 2014-03-26 21:23 Moon Kid <[hidden email]> wrote:
> Is this possible?

btw: Of course I found SIMPLAN but didn't unterstand it. I think
simulating data is quite simple task. variable, min, max, thats all.
SIMPLAN looks quite to complex for that.

=====================
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