coef(x) must be a two-way tensor or a matrix.
Since \(p\)-value depend on \(\widehat{\mathrm{cov}}^{-1}\{\mathrm{vec}(\mathbf{X})\}\) which is unavailable for the ultra-high dimensional \(\mathrm{vec}(\mathbf{X})\) in tensor predictor regression (TPR), the \(p\)-value plot is not provided for the object returned from TPR.fit.
Therefore, for the object return from TPR.fit, only the coefficients plot is displayed. And for the object return from TRR.fit, both the coefficients plot and \(p\)-value plot are displayed.
main and main_p control the titles of coefficient plot and \(p\)-value plot separately. Some other arguments used in function graphics::image, e.g., xlim, ylim, zlim, col, xaxs, yaxs, etc., can be passed to ...
ask can be set as FALSE if the pause before the second plot is not preferred. If x is an object from TPR.fit, no pause is enabled.