Page 1 of 1

Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 8:28 am
by berten
I have a dated panel for 51 cross-sections and 13 periods. I would like to add 20 more cross-sections, for which I have data in excel. This sounds trivial, but when I try to increase the cross-section ID series, EViews doesn't let me do that. How can I include these additional cross-sections? Thanks very much.

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 8:33 am
by EViews Gareth
If you use File->Import-Import From File, to import the Excel file, EViews should detect the panel nature of the incoming data, and the panel nature of the current workfile, and handle it automatically.

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 10:56 am
by berten
Thanks, I will try this. I have a different question as well. To calculate the normalized values of a variable, should I simply use @cnorm(x)? Would there be additional specifications for panel data? Thank you.

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 11:09 am
by EViews Gareth
Define "normalized".

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 11:48 am
by berten
I am also trying to figure out what normalized means in this context. I am replicating the results of a study which used an average of 12 normalized individual components to create an index. I attached one of these components. Their results lie between 0 and 1, mostly around 0.5 and 0.7. I assume this is what excel function norm.dist does. One concern is which mean and standard deviation is taken, is it for each country, or for all countries as a whole? We need the index to be comparable across countries.
I tried the above EViews command and it didn't work. It returned only 1 for each data point. How else can I do it?
Thanks.

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 12:11 pm
by EViews Gareth
There is no command to normalize. Just take off the mean and divide by the stdev.

Re: Adding new cross-sections to a dated panel

Posted: Wed Mar 20, 2013 12:17 pm
by berten
Sure, but doing that gives values greater than 1.