TFDEA ver 2.1 - Error message

[Saranya Durairajan]

Hello Dong-Joon,

Please find the attached excel. We have a few questions in this.

We tried to run our Spec 2000 dataset using frontier year 2005 and it ran successfully (Results 4Outputs sheet). But since the MAD was way too high, we plotted the SPEC trends and found processors in year 2006 could potentially impact our result. So, we decided to run the TFDEA model with frontier year as 2006 and then was the problem!

Please see "Results Frontier 2006" sheet. It shows a strange value "NaN" in MAD cell.

So, to cross check we wanted to try running Dr.A's dataset again and it runs successfully when the X-INVTECH (feature size) values are in decimals (actually it should be the reciprocal of it). So, when we tried to run after reciprocating, we get a value of '!' in MAD (please see the last sheet).

Now, the question is why are we getting strange values in MAD when we change the frontier date or when reciprocate a value? Also, why is TFDEA not forecasting anything after the frontier year in our case? Could you please help?

Thank you,

Saranya (on behalf of Team 4)