It's the first time I'm doing a response-surface analysis in R. I found the rsm package and was trying out their tutorial page when I came across this issue:
pacman:: p_load (dplyr, rsm)
table <- xlsx:: read.xlsx (file = "./Tabelas/fatorial_completo.xlsx", sheetIndex = 1, header = T) %>%
as.coded.data (x1~(PH-5)/1, x2~(NC-1000)/200)
model <- rsm (data = table, RP ~ SO (x1, x2))
summary (model)
Error in canonical(object, threshold = threshold) :
threshold is greater than the largest |eigenvalue|
In addition: Warning message:
In summary.lm(object = list(coefficients = c(`(Intercept)` = 100, :
essentially perfect fit: summary may be unreliable
Does anyone know why I'm getting this error message? I haven't been able to figure it out yet...
DATA
structure(list(x1 = c(-1, -1, -1, 0, 0, 0, 1, 1, 1),
x2 = c(-1, 0, 1, -1, 0, 1, -1, 0, 1),
PH = c(4, 4, 4, 5, 5, 5, 6, 6, 6),
NC = c(800, 1000, 1200, 800, 1000, 1200, 800, 1000, 1200),
RP = c(166.6666667, 100, 33.33333333, 166.6666667, 100, 33.33333333, 166.6666667, 100, 33.33333333)),
class = "data.frame", row.names = c(NA, -9L))
