Match Files Question

[Eugenio Grant]

Hi Guys:



I have to create a big DataBase from several files (file1, file2, file3,
file4, actually there are 33 files) using the MATCH FILES COMMAND.



Is there a way thru syntax to validate that every file (file1 to file4) has
the same ID sequence. So for example if file2 has less records, then the
process cannot be done. Only if all files have the same # of records then
the MATCH files should be done.



Any ideas???

[Frommelt, Allen]

We need to calculate 7 day and 14 day readmission rates for Medicare inpatients based on discharge date and the sequence of visit number (e.g., 2005-10-01, visit number=15) where each line represents one patient visit.  I used identify duplicate cases to determine the sequence number.  Does anyone know how to do this?

Thanks,

Allen

[Maguin, Eugene]

Eugenio,

Read up on the match files command. Although I have never done it, I think
you can do

Match files file=file1/in=f1/
   file=file2/in=f2/
   file=file3/in=f3/
   file=file4/in=f4/by key.

I'd expect the resulting file to have four variables, f1 to f4, with values
of 0 or 1. The main thing you are interested in, I think, is whether f1 thru
f4 have the same number of 1s.

Gene Maguin

[Maguin, Eugene]

Allen,

Probably others understand what you mean by 7 day and 14 day readmission
rates. I don't think I quite do understand what you mean. As I understand
what you said, you have mulitiple records per person id. You say you have a
sequence number and a discharge date. You must also have a person id. If you
only have discharge dates how do you know if they were readmitted? It would
seem to me that you need to have admission dates. So if a person were
discharged 40 days ago, and then again 20 days ago, how do you know whether
they were readmitted within 7 or 14 days of the discharge that occurred 40
days ago?

Gene Maguin

[Melissa Ives]

How are you defining the readmission rate?  By client only or by
admission?
Eg. Do you want to say something like:
A) Of the 200,000 admissions, 17% were readmissions within 7 days of
discharge and 23% were within 14 days of discharge.
B) Of the 200,000 admissions, 23% were readmissions within 7-14 days of
discharge from the previous admission.
C) Of the 200,000 admissions by 116,000 clients, 23% of clients had at
least one readmission within 7 days of discharge from the previous
admission.
D) etc..?

Melissa
The bubbling brook would lose its song if you removed the rocks.


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Frommelt, Allen
Sent: Tuesday, March 06, 2007 9:54 AM
To: [hidden email]
Subject: [SPSSX-L] Calculating Readmission Rates

We need to calculate 7 day and 14 day readmission rates for Medicare
inpatients based on discharge date and the sequence of visit number
(e.g., 2005-10-01, visit number=15) where each line represents one
patient visit.  I used identify duplicate cases to determine the
sequence number.  Does anyone know how to do this?

Thanks,

Allen


PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at [phone] for assistance.

[Melissa Ives]

I don't know how to stop the match from happening, but you could use the
/IN=varname subcommand to indicate which files contribute to the
dataset.  Then use select statements to see which ones don't match/keep
only those in all datasets.

MATCH FILES
        FILE=file1/IN=inf1/
        FILE=file2/IN=inf2/
        FILE=file3/IN=inf3/
        FILE=file4/IN=inf4/
        by Idvar.

SEL if inf1=1 and inf2=1 and inf3=1 and inf4=1.
*** or.
SEL if sum(inf1,inf2,inf3,inf4) eq 4.

*To see which ids are NOT in all files.
Temp.
SEL if sum(inf1,inf2,inf3,inf4) ne 4.
List vars=idvar inf1 inf2 inf3 inf4.

Melissa
The bubbling brook would lose its song if you removed the rocks.


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Eugenio Grant
Sent: Tuesday, March 06, 2007 9:27 AM
To: [hidden email]
Subject: [SPSSX-L] Match Files Question

Hi Guys:



I have to create a big DataBase from several files (file1, file2, file3,
file4, actually there are 33 files) using the MATCH FILES COMMAND.



Is there a way thru syntax to validate that every file (file1 to file4)
has the same ID sequence. So for example if file2 has less records, then
the process cannot be done. Only if all files have the same # of records
then the MATCH files should be done.



Any ideas???


PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at [phone] for assistance.

[Maguin, Eugene]

All,

Some questions about whether I understand the ordinal regression output. The
dv has three categories and the iv has two. The link is the logit. The
regression parallelism is satisfied. The results.

Parameter Estimates      Estimate
Threshold
        [A = 1] -.894
        [A = 2] .344

Location
        [C=0]           .593
        [C=1]           0(a)

First, what is the reference cateogry for the DV? The (syntax) documentation
says (of course) nothing. As well as the alogrithms manual and the case
studies thing. The advanced models manual says only that 'The ordering is
determined by sorting the values of the dependent variable in ascending
order. The lowest value defines the first category.' Now then, is the first
category labeled '0' or '1'? Is the first value the reference value (as is
true for logistic regression) or is it the last (highest value)? Does anyone
know?

Since the parallelism assumption is satisfied,
Can I say that relative to the dv reference category (whatever it is), the
odds of being in A=1 for C=0 is exponential(-.894+(.593-0))=.740?
And that the odds of being in A=2 for C=0 is
exponential(.344+(.593-0))=2.55?

Thanks, Gene Maguin

[Reutter, Alex]

Gene,

Ordinal regression doesn't have a reference category for the DV.

For ordinal regression models with the logit link, the cumulative odds ratio is generally reported.  The cumulative odds ratio at x1 vs x2 is simply exp(-B(x1-x2)) for DV categories k=1,...,K-1 (recall that P(Y<=K)=1 regardless of the value of the independents).

Cheers,
Alex


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Gene Maguin
Sent: Wednesday, March 07, 2007 9:41 AM
To: [hidden email]
Subject: [BULK] Computing odds ratios for ordinal regression
Importance: Low

All,

Some questions about whether I understand the ordinal regression output. The
dv has three categories and the iv has two. The link is the logit. The
regression parallelism is satisfied. The results.

Parameter Estimates      Estimate
Threshold
        [A = 1] -.894
        [A = 2] .344

Location
        [C=0]           .593
        [C=1]           0(a)

First, what is the reference cateogry for the DV? The (syntax) documentation
says (of course) nothing. As well as the alogrithms manual and the case
studies thing. The advanced models manual says only that 'The ordering is
determined by sorting the values of the dependent variable in ascending
order. The lowest value defines the first category.' Now then, is the first
category labeled '0' or '1'? Is the first value the reference value (as is
true for logistic regression) or is it the last (highest value)? Does anyone
know?

Since the parallelism assumption is satisfied,
Can I say that relative to the dv reference category (whatever it is), the
odds of being in A=1 for C=0 is exponential(-.894+(.593-0))=.740?
And that the odds of being in A=2 for C=0 is
exponential(.344+(.593-0))=2.55?

Thanks, Gene Maguin

[Maguin, Eugene]

All,

Is it possible to set the display width of variable values in the data
window via a syntax command. (I know it can be done in a couple of ways from
the either of the two data window screens. I know about that. I'm not
interested in that.) I didn't find anything in either the menu help or the
syntax documentation. Could have missed it. Did I?

Thanks, Gene Maguin

[Peck, Jon]

VARIABLE WIDTH varlist (some-value).

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Gene Maguin
Sent: Thursday, March 08, 2007 9:32 AM
To: [hidden email]
Subject: [SPSSX-L] Setting display width for variables via syntax

All,

Is it possible to set the display width of variable values in the data
window via a syntax command. (I know it can be done in a couple of ways from
the either of the two data window screens. I know about that. I'm not
interested in that.) I didn't find anything in either the menu help or the
syntax documentation. Could have missed it. Did I?

Thanks, Gene Maguin

[Oliver, Richard]

Have you tried the Variable Width command?

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Gene Maguin
Sent: Thursday, March 08, 2007 9:32 AM
To: [hidden email]
Subject: Setting display width for variables via syntax

All,

Is it possible to set the display width of variable values in the data
window via a syntax command. (I know it can be done in a couple of ways from
the either of the two data window screens. I know about that. I'm not
interested in that.) I didn't find anything in either the menu help or the
syntax documentation. Could have missed it. Did I?

Thanks, Gene Maguin

[ncanfield]

Hello!

When I import a table via the Open Database>New Query from an Access
table or Excel file, the date always defaults to the format of
dd-mmm-yyyy hh:mm:ss and I'm always having to switch the SPSS file it to
mm/dd/yy.  In the Access table and Excel files they are in the format of
mm/dd/yyyy.  Is this an option in SPSS or syntax/script that fixes this
in the opening of a new database?  Suggestions?

Thanks!


CONFIDENTIALITY NOTICE:  This communication and any files or attachments transmitted with it may contain information that is confidential, privileged and exempt from disclosure under applicable law.  It is intended solely for the use of the intended recipient.  If you are not the intended recipient, you are hereby notified that any unauthorized review, use, disclosure, dissemination, or copying of this communication is strictly prohibited.  If you have received this communication in error, please notify the sender by reply e-mail and destroy all copies of the original message.  Additionally, we will take the appropriate action to avoid sending you an unintended e-mail in the future.  Thank you for your cooperation.

[Albert-Jan Roskam]

Hi Gene (no pun intended ;-),

variable width yourvar (20).
and while you're at it, perhaps you also want:
variable alignment yourvar (center).

Cheers!!
Albert-Jan

--- Gene Maguin <[hidden email]> wrote:

> All,
>
> Is it possible to set the display width of variable
> values in the data
> window via a syntax command. (I know it can be done
> in a couple of ways from
> the either of the two data window screens. I know
> about that. I'm not
> interested in that.) I didn't find anything in
> either the menu help or the
> syntax documentation. Could have missed it. Did I?
>
> Thanks, Gene Maguin
>

Cheers!
Albert-Jan

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Did you know that 87.166253% of all statistics claim a precision of results that is not justified by the method employed? [HELMUT RICHTER]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



____________________________________________________________________________________
Never miss an email again!
Yahoo! Toolbar alerts you the instant new Mail arrives.
http://tools.search.yahoo.com/toolbar/features/mail/