Home>

built a linear model comprising a dependent continuous variable, 3 continuous predictors, 2 discrete (factors with two gradations each), as well as the interaction of one of these continuous and one discrete predictor. I want to build a series of a series of graphs, reflecting the connection of the dependent variable with predictors, as well as on one of them to visualize the interaction. The course was taught to depict the connection with the most important predictors, provided that all other predictors are unchanged, for example, their values ​​are equal to average. Only how to make the factors unchanged? Averaging them, understandable, will not work equal to zero -also does not accept.

  • Answer # 1

    The question is quite common, therefore the answer is the same.

    Simple and unambiguous way to assess the relationship of a single predictor with a target variable within a particular model does not exist. There are a number of approaches that are discussed in detail in the book. https://ema.drwhy.ai/(In particular, see Break-Down Plots and Partial-Dependence Profiles).

  • Answer # 2

    The question is quite common, therefore the answer is the same.

    Simple and unambiguous way to assess the relationship of a single predictor with a target variable within a particular model does not exist. There are a number of approaches that are discussed in detail in the book. https://ema.drwhy.ai/(In particular, see Break-Down Plots and Partial-Dependence Profiles).