Technical efficiency analysis

[Ilya Kozyr]

Hello. I have a problem when I try to analyze technical efficiency.

I have no problems when I use code:
xtfrontier ..., ti
predict te,te

But when I use tvd model and try to predic te, I see error r133() "unknown egen function max()"

Can ypu help, what is a problem and how I can calculate te for tvd model?
Thanks

[eric_a_booth]

You'll have to explain more about what your data look like and your syntax you are running. This example runs without error:

webuse xtfrontier1,clear
 xtfrontier lnwidgets lnworkers lnmachines , ti  
predict te,te
 xtfrontier lnwidgets lnworkers lnmachines , tvd  
 predict te2,te

[Ilya Kozyr]

I use panel data
My code looks like:

xtset bankid qtr
xtfrontier roe loans sec pdep pfixas crrisk liqrisk caprisk as as_sq gov_new for_new federal other_center res_1, ti
predict te, te

xtfrontier roe loans sec pdep pfixas crrisk liqrisk caprisk as as_sq gov_new for_new federal other_center res_1, tvd
predict te1, te
unknown egen function max()
r(133)

So technical efficiency calculated for time-invariant model but there is an error for time-varian

[eric_a_booth]

In the background, predict is running a command like:

by bankid: egen double var = max(qtr)

Which gives the error, but I've never seen Stata error on this egen function. What version of Stata are you using; is it up to date? Finally, try running the egen function 'max' outside of -predict- to see if it's a problem with -predict- after xtfrontier or a problem with -egen-.

[Ilya Kozyr]

I use Stata13

I think that there is a problem with -egen-, because for time-inveriant model command -predict- works perfect. Actually a can male no operation with -egen- command

Moreover, I try to install a packege

. ssc install egenmore
checking egenmore consistency and verifying not already installed...
all files already exist and are up to date.

[Ilya Kozyr]

Actually I even try to use manual about -egen-, so I use special data and code which was written there and

use http://www.stata-press.com/data/r13/egenxmpl2, clear
by dcode, sort: egen medstay = median(los)
generate deltalos = los - medstay

But my Stata again face wwith errors


. use http://www.stata-press.com/data/r13/egenxmpl2, clear

.
. by dcode, sort: egen medstay = median(los)
unknown egen function median()
r(133);

.
. generate deltalos = los - medstay
medstay not found
r(111);

[eric_a_booth]

The max and median egen functions are not a part of the SSC package -egenmore- (maybe you are thinking of the function 'rowmedian' which at one time originated from egenmore? ). Regardless, if a canned example like

sysuse auto, clear

bys rep78: egen a = max(price)
bys rep78: egen b = median(price)

is failing then my recommendations are to reinstall Stata 13 and/or contact Stata Technical support.

[Ilya Kozyr]

Many thanks for your advices.
It`s again error, I write to support