[Limdep Nlogit List] Problem with panel data

[dgiantsios at uom.gr]

Hello,

Iam using a balanced panel for Battese-Coelli Stochastic Frontier Analysis
with nine observations for each group (Pds=9)but there exist missing
values. Unfortunatelly,some missing values are at the beggining, middle or
end of each group. When i define Global Panel settings, while i have max
group size=9 in my sample, in the results the program seems like this with
maximum group size=18:
 Variable = ____________ Variable Groups    Max    Min   Average |
| TI         Group sizes  ID          946     18      3       8.4 |
+-----------------------------------------------------------------+
+------------------------------------------------------+
| Frequency count for group sizes of TI                |
| Group size =   3   Pct =    .74%   CumPct =    .74%  |
| Group size =   4   Pct =   1.16%   CumPct =   1.90%  |
| Group size =   5   Pct =   1.06%   CumPct =   2.96%  |
| Group size =   6   Pct =   3.07%   CumPct =   6.03%  |
| Group size =   7   Pct =   7.72%   CumPct =  13.74%  |
| Group size =   8   Pct =  22.30%   CumPct =  36.05%  |
| Group size =   9   Pct =  63.85%   CumPct =  99.89%  |
| Group size =  10   Pct =    .00%   CumPct =  99.89%  |
| Group size =  11   Pct =    .00%   CumPct =  99.89%  |
| Group size =  12   Pct =    .00%   CumPct =  99.89%  |
| Group size =  13   Pct =    .00%   CumPct =  99.89%  |
| Group size =  14   Pct =    .00%   CumPct =  99.89%  |
| Group size =  15   Pct =    .00%   CumPct =  99.89%  |
| Group size =  16   Pct =    .00%   CumPct =  99.89%  |
| Group size =  17   Pct =    .00%   CumPct =  99.89%  |
| Group size =  18   Pct =    .11%   CumPct = 100.00%

How i must define the panel but the SKIP comand will not change my group
sie???

Best Regards
Dimitrios