svmInference: l1-SVR regression rankscore test
In jkcshea/l1svr: Inference for Support Vector Regression Under l1 Regularization

This function performs the l1-SVR regression rankscore test for a single coefficient, and returns the p-value.

1
2
3

svmInference(Xr, Zr, Y, U, epsilon, lambda = 0, solver = "gurobi",
  intercept = TRUE, nullGamma = 0, heteroskedastic = FALSE,
  hc = 0.5, kappa = 1)

`Xr`	Matrix of covariates, excluding the covariate of interest for which inference is being conducted.
`Zr`	Vector of the covariate of interest for which inference is being conducted.
`Y`	Vector of the outcome.
`U`	Vector of the residuals.
`epsilon`	Real scalar, bandwidth for SVR.
`lambda`	Real scalar, parameter for SVR that scales the penalty.
`solver`	Character, name of the linear programming package in R used to obtain the bounds on the treatment effect. The function supports `'gurobi'`, `'cplexapi'`, `'lpsolveapi'`. The name of the solver should be provided with quotation marks.
`intercept`	Boolean, set to `TRUE` if an intercept should be included in the regression.
`nullGamma`	Real scalar, the coefficient on `Zr` under the null.
`heteroskedastic`	Boolean, indicate whether data has heteroskedastic errors. If set to `TRUE`, then density estimation is performed when conducting inference.
`hc`	Scalar determing the quantiles of a standard normal distribution used to define the bandwidth for nonparametric density estimation. See Powell (1991) and Bai et. al. (2021) for details.
`kappa`	Scalar that directly scales the bandwdith for nonparametric density estimation. See Powell (1991) and Bai et. al. (2021) for details.