Prev: ERRORABEND vs INVALID DATA LINES
Next: How to Make Sauce for dry dog food?
From: Church on 8 Feb 2010 05:38
Hi,

I have a time series and the projected values (horizon: 50 years)
should be greater then zero. Therefore I want to include a log
transformation. However, the proc arima statement gives no possibility
to do such a transformation.
How can I achieve this, preferably in a standardized way?

Thanks.
From: Wensui Liu on 9 Feb 2010 23:17
could you take the log transformation in data step?

On Mon, Feb 8, 2010 at 5:38 AM, Church <H.J.Plat(a)uva.nl> wrote:
> Hi,
>
> I have a time series and the projected values (horizon: 50 years)
> should be greater then zero. Therefore I want to include a log
> transformation. However, the proc arima statement gives no possibility
> to do such a transformation.
> How can I achieve this, preferably in a standardized way?
>
> Thanks.
>



--
==============================
WenSui Liu
Blog : statcompute.spaces.live.com
Tough Times Never Last. But Tough People Do. - Robert Schuller
==============================
 | 
Pages: 1
Prev: ERRORABEND vs INVALID DATA LINES
Next: How to Make Sauce for dry dog food?