Converting timeseries to Panel Data

[amirkh]

I run a pilot project which will eventually lead to a large scale of data management in the next phrase of the project. I need to convert standard reports from a database to panel data.
The reports are retrieved from the database in the following format. In the real case I will have more than 20 variables (e.g. I/GL) and 6 years of data for more than 500 banks.

"Bank Name" "I/GL 2009" "I/GL 2008" "I/GL 2007"
1. Federal National 6,30 1,60 0,80
2. HSBC Holdings Plc 3,30 2,70 1,80
3. Barclays Bank Plc 5,20 3,40 2,50
4. Barclays Plc 5,20 3,40 2,80

My preferred format (Panel Data) would be:

"Bank Name" "Year" "I/GL"
1. Federal National 2007 0,8
1. Federal National 2008 1,60
1. Federal National 2009 6,3
2. HSBC Holdings Plc 2007 1,8
2. HSBC Holdings Plc 2008 2,7
2. HSBC Holdings Plc 2009 3,3
3. Barclays Bank Plc 2007 2,50
3. Barclays Bank Plc 2008 3,4
3. Barclays Bank Plc 2009 5,2
4. Barclays Plc 2007 2,8
4. Barclays Plc 2008 3,4
4. Barclays Plc 2009 5,2

Can any one help me to do this conversion?

Many thanks in advance,
Amir

[EViews Gareth]

Have you read:
http://forums.eviews.com/viewtopic.php?f=7&t=74

[amirkh]

Thanks for the response and the answer. I checked the link you sent me before. The data constructed differently still from those cases presented by the link you sent me. Still have problem since need to seperate both, year and variable.

I here attach a file which might help you understand it better.

[EViews Gareth]

Stack it using "il_?" as the stacking identifier. Then restructure the page to use bank_name as the cross-section ID and VAR01 as the date id.