.X12 seasonal adjustment function

[eviewsnew]

Hi,

I don't know how this .X12 seasonal adjustment works. Basically, in implementation, 1. does it matter if the data frequency is quarterly or mothly or annual? do I have to have a series to indicate its frequency?

2. if I have a 10*100 matrix, with each of the 10 rows representing a region and 100 quarterly time series in 100 collumns. Can I use .X12 function to let it do seasonal adjustment for each row? if how does it detect the time frequency if it's a matrix but not series? Or I have to transpose the matrix into series with 1000 rows and one collumn for the time series and another collumn to indicate its frequency (2010Q1 pattern) before using the .X12 function?

thanks!

[trubador]

Seasonal adjustment works on monthly or quarterly data. The data should not be in matrix/vector format. If that is the case, then you should convert the data into series form (look for "mtos" function). Your workfile should be structured as monthly or quarterly, since this is how X12 procedure detects the time frequency. Please refer to users guide for details...

[eviewsnew]

Got your point, thanks!

But a data transformation question here:

If I have this dataset from excel in 10 rows with each row as a region having 100 collumns of econ data for each quarter starting from 1980Q1, after I import it into Eviews, how can I transform it to the series you mentioned for the .X12?

What I think is: get the data in, and transpose it to 1000*2 with each 100 rows for 1 region. And then the collumn row has the econ data. With this new matrix I can do stom to make series I may be able to do the .X12 function with loop. But how to create a seperate series for the frequency information - 1980Q1 ---- to match this new series data?

thanks a lot!

[EViews Gareth]

Probably easier just to transpose it in Excel, then open it in EViews.

[eviewsnew]

I c. I'm doing it in SAS.

Another question, when I export series with WFSAVE, if the series names are area_0, area_1, ....area_300. And I need to export only the first 256 series to one .csv file (limit) and the rest into a second group, how do I name those series in abbreviation but not listing all of them?

[EViews Gareth]

wfsave has a @keep argument that lets you specify which series to keep. The documentation has more details.