The
solder data frame, a design object, contains 900 observations
(rows), which are the results an
experiment varying 5
factors relevant to the wave-soldering procedure for mounting
components on printed circuit boards. The response variable,
skips
is a count of how many solder skips appeared to a visual inspection.
The
solder.balance data frame,
contains 720 rows representing a balanced subset of
all the experimental runs. The variables (columns) are the same; it is
also a design object.
The
solder2 contains 750 observations which are the results of a similar
experiment incorporating some of the knowledge gained in the first.
The factors still have the same names and it is still a design object.
These data frames contains the following columns:
A1.5 ,
A3 ,
A6 ,
B3 , and
B6 .
W4 ,
D4 ,
L4 ,
D6 ,
L6 ,
D7 ,
L7 ,
L8 ,
W9 , and
L9 .
1 ,
2 or
3 on a board is being counted.
Comizzoli, R. B, Landwehr, J. M., and Sinclair, J. D. (1990)
Robust Materials and Processes:
Key to Reliability. AT&T Technical Journal Vol. 69, No. 6, pp. 113--128.
John M. Chambers and Trevor J. Hastie, Statistical Models in S,
Wadsworth and Brooks, Pacific Grove, CA 1992, pg. 7.
fit1 <- aov(sqrt(skips) ~ ., data = solder.balance) fit2 <- aov(sqrt(skips) ~ .^2, data = solder.balance)