This was the first time I needed to look at the confidence intervals of the time totals, so I used the option level=0.95 with the totals() command. To my surprise, the output I got was really strange: the values in the lo and hi columns were identical and marginally below the 'imputed' value.
The same strange output I got also using the package's own 'skylark' data instead of my data:
data(skylark)
z <- trim(count ~ site + time, data=skylark, model=3)
totals(z, level=0.95)
time imputed se_imp lo hi
1 1 511 38 510.0584 510.0584
2 2 362 31 361.1155 361.1155
3 3 429 26 428.4749 428.4749
4 4 423 25 422.5076 422.5076
5 5 469 27 468.4820 468.4820
6 6 522 27 521.5346 521.5346
7 7 562 32 561.3928 561.3928
8 8 606 36 605.2873 605.2873
The rtrim version I used was the latest available 2.3.0.
best wishes
Ainars Aunins
This was the first time I needed to look at the confidence intervals of the time totals, so I used the option level=0.95 with the totals() command. To my surprise, the output I got was really strange: the values in the lo and hi columns were identical and marginally below the 'imputed' value.
The same strange output I got also using the package's own 'skylark' data instead of my data:
The rtrim version I used was the latest available 2.3.0.
best wishes
Ainars Aunins