Description Usage Arguments Value Examples
The consideration set model takes data with the format of each row representing a choice for some person. First, the model constructs a set of choices in consideration for each person via a binary logit model (the Selection step), then selects choices from consideration set via a multinomial logit model (the Consideration step).
1 2 3 | RunConsiderationSet(YObservedInput, XConsidInput, XSelectInput, NumOpts,
ASCConsid = 0, ASCSelect = 0, IncludedInput = 0 * YObservedInput + 1,
meth = "L-BFGS-B", Tolerance = 1e-05)
|
YObservedInput |
A vector of observed choices acting as the response variable for all subjects, each row representing a unique choice and the value representing how many times it was chosen. |
XConsidInput |
A vector or matrix of factors that influences what choices are to be chosen from the consideration set, i.e. for second step. |
XSelectInput |
A vector or matrix of factors that influences what choices are to be included in the consideration set, i.e. for first step. |
NumOpts |
Number of options for consideration set. Due to the nature of the model, NumOpts needs to be a factor of the total number of rows, such that every n (value of NumOpts) rows represent a set of observed choices for one subject. |
ASCConsid |
Binary flag to indicate if users want ASC, i.e. intercept in the Consideration step; 0, i.e. "No", by default. |
ASCSelect |
Binary flag to indicate if users want ASC, i.e. intercept in the Selection step; 0, i.e. "No", by default. |
IncludedInput |
A vector or matrix of which choices are available for constructing the model; every choice is available, i.e. a vector of 1's for all YObservedInput, by default. (A value of 0 excludes a particular choice.) |
meth |
Algorithm used as method for optimization. Methods available are "L-BFGS-B", "BFGS", "CG", "SANN", "Brent"; "L-BFGS-B" by default. |
Tolerance |
(Valid only when |
A list object containing the following:
Likelihood for the consideration set model.
Likelihood for the logit model.
Coefficients for all covariates in Consideration and Selection input. If ASCConsid / ASCSelect is set equal to 1, the output will include ASC in Consideration / Selection.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | data(CFS5-1)
attach(CFS5-1)
RunConsiderationSet(V1, cbind(V2,V3,V4), cbind(V2,V3,V4), 5, 1, 1,Tolerance=1e-6)
data(CFS6-1)
attach(CFS6-1)
RunConsiderationSet(V1, cbind(V2,V3,V4), cbind(V2,V3,V4), 6, 1, 1)
data(CFS5-2)
attach(CFS5-2)
RunConsiderationSet(V1, cbind(V2,V3,V4), cbind(V2,V3,V4), 5, 1, 1, IncludedInput=V5, meth="BFGS")
data(CFS6-2)
attach(CFS6-2)
RunConsiderationSet(V1, cbind(V2,V3,V4), cbind(V2,V3,V4), 6, 1, 1, IncludedInput=V5, meth="BFGS")
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.