The case study is described in several sources:
Here are the data, in standard order:
Standard order | A | B | C | D | y1 | y2 |
1 | -1 | -1 | -1 | -1 | 43.5 | 82 |
2 | 1 | -1 | -1 | -1 | 51.3 | 83.7 |
3 | -1 | 1 | -1 | -1 | 35 | 61.7 |
4 | 1 | 1 | -1 | -1 | 38.4 | 100 |
5 | -1 | -1 | 1 | -1 | 44.9 | 82.1 |
6 | 1 | -1 | 1 | -1 | 52.4 | 84.1 |
7 | -1 | 1 | 1 | -1 | 39.7 | 67.7 |
8 | 1 | 1 | 1 | -1 | 41.3 | 100 |
9 | -1 | -1 | -1 | 1 | 41.3 | 82 |
10 | 1 | -1 | -1 | 1 | 50.2 | 86.3 |
11 | -1 | 1 | -1 | 1 | 37.5 | 66 |
12 | 1 | 1 | -1 | 1 | 39.2 | 100 |
13 | -1 | -1 | 1 | 1 | 43 | 82.2 |
14 | 1 | -1 | 1 | 1 | 51.9 | 89.8 |
15 | -1 | 1 | 1 | 1 | 39.9 | 68.6 |
16 | 1 | 1 | 1 | 1 | 41.6 | 100 |
And the source code to use in R (copy and paste below; or, run the code in a web-browser)
# Solar panel case study, from BHH2, p 230 install.packages("pid") # Only required once, if you've not installed "pid" already library(pid) data("solar") model_y1 <- lm(y1 ~ A*B*C*D, data=solar) # collection efficiency model summary(model_y1) paretoPlot(model_y1) model_y2 <- lm(y2 ~ A*B*C*D, data=solar) # energy delivery efficiency model summary(model_y2) paretoPlot(model_y2) |