merge with devel

NEWS.md

@@ -18,6 +18,10 @@ As of April 2021:
 * Support for the custom lasso procedure implemented in `Rcpp` has been
   discontinued. Accordingly, the `"lassi"` option and argument `fit_type` have
   been removed from `fit_hal`.
+* Re-added `lambda.min.ratio` as a `fit_control` argument to `fit_hal`. We've
+  seen that not setting `lambda.min.ratio` in `glmnet` can lead to no `lambda`
+  values that fit the data sufficiently well, so it seems appropriate to
+  override the `glmnet` default.
 
 # hal9001 0.3.0
 

R/hal.R

@@ -130,6 +130,11 @@
 #'   \code{\link[glmnet]{cv.glmnet}}. When \code{TRUE}, \code{"lambda.min"} is
 #'   used; otherwise, \code{"lambda.1se"}. Only used when
 #'   \code{cv_select = TRUE}.
+#' - \code{lambda.min.ratio}: A \code{\link[glmnet]{glmnet}} argument specifying
+#'   the smallest value for \code{lambda}, as a fraction of \code{lambda.max},
+#'   the (data derived) entry value (i.e. the smallest value for which all
+#'   coefficients are zero). We've seen that not setting \code{lambda.min.ratio}
+#'   can lead to no \code{lambda} values that fit the data sufficiently well.
 #' - \code{prediction_bounds}: A vector of size two that provides the lower and
 #'   upper bounds for predictions. When \code{prediction_bounds = "default"},
 #'   the predictions are bounded between \code{min(Y) - sd(Y)} and
@@ -173,8 +178,8 @@ fit_hal <- function(X,
                     X_unpenalized = NULL,
                     max_degree = ifelse(ncol(X) >= 20, 2, 3),
                     smoothness_orders = 1,
-                    num_knots = sapply(seq_len(max_degree),
-                      num_knots_generator,
+                    num_knots = num_knots_generator(
+                      max_degree = max_degree,
                       smoothness_orders = smoothness_orders,
                       base_num_knots_0 = 500,
                       base_num_knots_1 = 200
@@ -189,6 +194,7 @@ fit_hal <- function(X,
                       n_folds = 10,
                       foldid = NULL,
                       use_min = TRUE,
+                      lambda.min.ratio = 1e-4,
                       prediction_bounds = "default"
                     ),
                     formula_control = list(
@@ -206,7 +212,7 @@ fit_hal <- function(X,
   # errors when a supplied control list is missing arguments
   defaults <- list(
     cv_select = TRUE, n_folds = 10, foldid = NULL, use_min = TRUE,
-    prediction_bounds = "default"
+    lambda.min.ratio = 1e-4, prediction_bounds = "default"
   )
   if (any(!names(defaults) %in% names(fit_control))) {
     fit_control <- c(
@@ -472,12 +478,16 @@ fit_hal <- function(X,
 #'  the basis function.
 #'
 #' @keywords internal
-num_knots_generator <- function(d, smoothness_orders, base_num_knots_0 = 500,
+num_knots_generator <- function(max_degree, smoothness_orders, base_num_knots_0 = 500,
                                 base_num_knots_1 = 200) {
   if (all(smoothness_orders > 0)) {
-    return(round(base_num_knots_1 / 2^(d - 1)))
+    return(sapply(seq_len(max_degree), function(d) {
+      round(base_num_knots_1 / 2^(d - 1))
+    }))
   }
   else {
-    return(round(base_num_knots_0 / 2^(d - 1)))
+    return(sapply(seq_len(max_degree), function(d) {
+      round(base_num_knots_0 / 2^(d - 1))
+    }))
   }
 }

R/summary.R

@@ -62,7 +62,7 @@ summary.hal9001 <- function(object,
     if (lambda != object$lambda_star) {
       if (is.null(object$lasso_fit)) {
         stop(
-          "Coefficients for specified lamdba do not exist, or are not ",
+          "Coefficients for specified lambda do not exist, or are not ",
           "accessible since the fit of the lasso model was not returned ",
           "(i.e., return_lasso was set to FALSE in `hal_fit()`)."
         )
@@ -85,7 +85,7 @@ summary.hal9001 <- function(object,
     coefs <- object$coefs
     if (length(lambda) > 1) {
       warning(
-        "Coefficients for many lamdba exist --\n",
+        "Coefficients for many lambda exist --\n",
         "Summarizing coefficients corresponding to minimum lambda."
       )
       lambda_idx <- which.min(lambda)

README.Rmd

@@ -21,6 +21,7 @@ knitr::opts_chunk$set(
 [![Coverage Status](https://img.shields.io/codecov/c/github/tlverse/hal9001/master.svg)](https://codecov.io/github/tlverse/hal9001?branch=master)
 [![CRAN](https://www.r-pkg.org/badges/version/hal9001)](https://www.r-pkg.org/pkg/hal9001)
 [![CRAN downloads](https://cranlogs.r-pkg.org/badges/hal9001)](https://CRAN.R-project.org/package=hal9001)
+[![CRAN total downloads](http://cranlogs.r-pkg.org/badges/grand-total/hal9001)](https://CRAN.R-project.org/package=hal9001)
 [![Project Status: Active – The project has reached a stable, usable state and is being actively developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
 [![License: GPL v3](https://img.shields.io/badge/License-GPL%20v3-blue.svg)](http://www.gnu.org/licenses/gpl-3.0)
 [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.3558313.svg)](https://doi.org/10.5281/zenodo.3558313)

README.md

@@ -12,6 +12,8 @@ Status](https://img.shields.io/codecov/c/github/tlverse/hal9001/master.svg)](htt
 [![CRAN](https://www.r-pkg.org/badges/version/hal9001)](https://www.r-pkg.org/pkg/hal9001)
 [![CRAN
 downloads](https://cranlogs.r-pkg.org/badges/hal9001)](https://CRAN.R-project.org/package=hal9001)
+[![CRAN total
+downloads](http://cranlogs.r-pkg.org/badges/grand-total/hal9001)](https://CRAN.R-project.org/package=hal9001)
 [![Project Status: Active – The project has reached a stable, usable
 state and is being actively
 developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
@@ -111,7 +113,7 @@ hal_fit$times
 # training sample prediction
 preds <- predict(hal_fit, new_data = x)
 mean(hal_mse <- (preds - y)^2)
-#> [1] 0.0357806
+#> [1] 0.03481173
 ```
 
 -----
@@ -167,7 +169,7 @@ See file `LICENSE` for details.
 
 ## References
 
-<div id="refs" class="references hanging-indent">
+<div id="refs" class="references">
 
 <div id="ref-benkeser2016hal">