Importing data from Troll database

[Olister]

Hi,

I am currently using two softwares: Eviews and TROLL. Previously, the READ command could import only “ASCII”, “wk1”, or “xls” files type. Now with Eviews 7, by using of IMPORT command further files types are supported.

In the Option Keywords of the IMPORT command it is written as part of different file types:
RATS Portable / TROLL "l", "trl"

I tried to use these options, but it is not working. Here is some of the codes that I have tested:

import basetroll.trl
import(t=trl) basetroll.trl
import(t=trl) v:\mypath\basetroll.trl

The Error Message I get is: “Unsupported source format for import in “IMPORT BASE.TRL

The other Eviews commands, DBOPEN, WFOPEN, and DBCOPY, have RATS Portable / TROLL files types in their Option Keywords.

When I use DBCOPY: dbcopy v:\mypath\basetroll.trl baseviews.edb Eviews crashes and closes…

If I do not program my import but do it by hand through my workfile : Proc/import/Import from file…
In the combo box of the files types, there is not TROLL but only “Rats Portable – as workfile (*.trl)” available.

Am I using these commands correctly? What can I do to import data from TROLL? Is it possible?

Thanks for your help,

[EViews Gareth]

I don't think we've seen a TROLL file for a decade or so. Is there any chance you could send us the file that is causing the problems? support@eviews.com (if it is small enough for email).

[Olister]

Thanks Gareth, I just sent a TROLL database files to you.
I hope that will help you to help me :)

Olister

[johansamuelsson]

How's it going with this issue, we are also interested of transport data from TROLL to Eviews?

Regards Johan

[EViews Gareth]

Yep, it turns out that TROLL changed their format at some point in the last 15 years. We cannot read the new format. We've removed mention of TROLL from the documentation for EViews 8. I don't foresee introducing support for the new format.

I assume TROLL is able to save data in other formats (CSV? - everyone does CSV!), so you'll have to use that route.

[johansamuelsson]

OK, we try CSV