merging

[Greg]

Hi everyone,

I'm trying to merge two datasets, a sample which look like:

Dataset A Dataset B
900100 900100
900100 900100
900100 900100
900100 900200
900100 900200
900100 900300
900100 900300
900200 900300
900200 900300
900300 900400
900300 900400
900400 900400
                       
I want to match dataset B with dataset A. The datasets correspond to census designated PUMAS in 2000 (dataset A) and 2010 (dataset B). The pumas changed from 2000 to 2010. For example, the 7 '900100' pumas in 2000 became 3 '900100s' in 2010.

Any advice/tips are greatly appreciated and thank you!

Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[e-mail] (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

[Maguin, Eugene]

So, looking at your two lists the naïve person sees 7 900100 records in A and three in B. And since she's supposed to match these up she wants to know which, if any, B records go with which, if any, A records. What do you tell her?
Gene Maguin

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Greg
Sent: Tuesday, March 6, 2018 8:34 AM
To: [hidden email]
Subject: merging

Hi everyone,

I'm trying to merge two datasets, a sample which look like:

Dataset A Dataset B
900100 900100
900100 900100
900100 900100
900100 900200
900100 900200
900100 900300
900100 900300
900200 900300
900200 900300
900300 900400
900300 900400
900400 900400
                       
I want to match dataset B with dataset A. Any advice/tips are greatly appreciated and thank you!

Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Greg]

Hi Gene,

Thank you for your response. I edited my original post by adding some
(brief) context to the datasets.

Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Maguin, Eugene]

Greg, I don't have any experience with PUMA or PUMS files. I have heard of them because somebody here worked with American Community Survey data. When I googled PUMAs, one response was cartographic boundary shapefiles for public use from the census bureau (https://www.census.gov > Geography > Maps & Data > Cartographic Boundary Files). I wonder if, using ARC, you could put the 2000 shapefiles side by side with the 2010 shapefiles and get something useful enough to go forward.
Gene Maguin

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Greg
Sent: Tuesday, March 6, 2018 9:04 AM
To: [hidden email]
Subject: Re: merging

Hi Gene,

Thank you for your response. I edited my original post by adding some
(brief) context to the datasets.

Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Andy W]

The term to search for is "crosswalk". Basically you will do an areal
allocation to go from one set to another.

https://usa.ipums.org/usa/volii/puma00_puma10_crosswalk_spatial.shtml



-----
Andy W
[hidden email]
http://andrewpwheeler.wordpress.com/
--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Greg]

Ok and thank you. Though I'm not completely confident in working with GIS,
this cannot be done via spss merging?


Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Andy W]

No you can do this all within SPSS -- the link I provided has an excel
spreadsheet you can read into SPSS.

In a nutshell, you will need to aggregate the 2000 database based on the
proportions in that spreadsheet, then it will be a simple 1 to 1 merge.



-----
Andy W
[hidden email]
http://andrewpwheeler.wordpress.com/
--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Greg]

and this is what confuses me. So I aggregate based on the 2000 pumas (using
them as the break variable), but since some 2000 pumas split into more than
one puma in 2010, I will still have to do a 'many to many' merge? (I'm sure
I'm missing something(?).


Greg



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Andy W]

No, the crosswalk file should have both 2000 and 2010 identifiers. It will
then have a field that is a proportion. You will want to aggregate whatever
2000 information to the 2010 identifiers, using that proportion as a weight.

That is about as much specifics as I can give without more information about
your original data, to which I could construct an example with code.



-----
Andy W
[hidden email]
http://andrewpwheeler.wordpress.com/
--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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