I fit parasitoid abundance to two factors, flower availability and distance of the crop to flower.
Firstly, I fit separate glm equation:
glm(abundance ~ flower, data = data, family = poisson(link"log"))
which in this case, flower had two classes: "yes" and "no" the class "no" here regarded as control.
Then I fit abundance ~ distance
The data$distance had three classes (0,10,
and 20), but in dataframe, this column contains additional classes from data$flower= "no" .
Consequently, I eliminated the rows containing flower = "No" (or control) by subsetting it, because it not make sense for me to include the distance observation from control data,
glm(abundance ~ distance, data = data, family = poisson(link"log"),
subset = flower != "No")
The problem arises when I try to merge both factors into a single glm equation:
glm(abundance ~ flower + distance, data = data, family = poisson(link"log"))
how to pass the subset = flower != "No" , only to the variable data$distance without disturb other factor?
any suggestion?