Merge pull request #114 from s3alfisc/dev

v0.14

.github/workflows/R-CMD-check.yaml

@@ -81,7 +81,7 @@ jobs:
 
       # install WildBootTests.jl
       - name: install WildBootTests.jl 
-        run: julia -e 'using Pkg; Pkg.add(["WildBootTests", "StableRNGs"])'
+        run: julia -e 'using Pkg; Pkg.add(["StableRNGs"]); Pkg.add(Pkg.PackageSpec(;name="WildBootTests", version="0.9.9"))'
         # use shell bash to ensure consistent behavior across OS
         shell: bash
 

DESCRIPTION

@@ -1,9 +1,19 @@
 Package: fwildclusterboot
 Title: Fast Wild Cluster Bootstrap Inference for Linear Models
-Version: 0.13.2
+Version: 0.14.0
 Authors@R: c(
     person("Alexander", "Fischer", , "[email protected]", role = c("aut", "cre")),
     person("David", "Roodman", role = "aut"),
+    person("Megha", "Joshi", 
+            role = "rev", 
+            comment = "Megha reviewed the package (v. 0.13) for ropensci 
+                      , see <https://github.com/ropensci/software-review/issues/546>"
+          ),
+    person("Eunseop", "Kim",
+            role = "rev", 
+            comment = "Eunseop reviewed the package (v. 0.13) for ropensci 
+                      , see <https://github.com/ropensci/software-review/issues/546>"
+          ),
     person("Achim", "Zeileis", role = "ctb",
            comment = "Author of included sandwich fragments"),
     person("Nathaniel", "Graham", role = "ctb",
@@ -42,7 +52,6 @@ Imports:
     generics,
     gtools,
     JuliaConnectoR,
-    MASS,
     Matrix,
     Rcpp,
     rlang,
@@ -64,7 +73,8 @@ Suggests:
     rmarkdown,
     sandwich,
     testthat (>= 3.0.0),
-    tibble
+    tibble, 
+    MASS
 LinkingTo:
     Rcpp,
     RcppArmadillo,
@@ -77,9 +87,9 @@ Language: en-US
 LazyData: true
 Roxygen: list(markdown = TRUE, roclets = c ("namespace", "rd",
     "srr::srr_stats_roclet"))
-RoxygenNote: 7.2.1
+RoxygenNote: 7.2.3
 SystemRequirements: Version Requirements to run the wild bootstrap through
-    Julia - Julia (>= 1.8), WildBootTests.jl (>=0.9).  Julia is
+    Julia - Julia (>= 1.8), WildBootTests.jl (>=0.9.8).  Julia is
     downloadable via the official Julia website
     (https://julialang.org/downloads/), WildBootTests.jl via Julia's
     package manager (https://docs.julialang.org/en/v1/stdlib/Pkg/) or its

NAMESPACE

@@ -13,8 +13,6 @@ S3method(mboottest,lm)
 S3method(nobs,boottest)
 S3method(nobs,mboottest)
 S3method(plot,boottest)
-S3method(print,boottest)
-S3method(print,mboottest)
 S3method(pval,boottest)
 S3method(pval,mboottest)
 S3method(summary,boottest)
@@ -34,12 +32,18 @@ export(teststat)
 export(tidy)
 importFrom(JuliaConnectoR,juliaEval)
 importFrom(JuliaConnectoR,juliaImport)
-importFrom(MASS,ginv)
 importFrom(Matrix,Diagonal)
+importFrom(Matrix,Matrix)
+importFrom(Matrix,crossprod)
+importFrom(Matrix,sparse.model.matrix)
+importFrom(Matrix,t)
+importFrom(Matrix,tcrossprod)
 importFrom(Rcpp,evalCpp)
 importFrom(collapse,GRP)
+importFrom(collapse,fmean)
 importFrom(collapse,fsum)
 importFrom(collapse,qF)
+importFrom(collapse,qtab)
 importFrom(dqrng,dqsample)
 importFrom(dqrng,dqset.seed)
 importFrom(dreamerr,check_arg)

NEWS.md

@@ -1,3 +1,37 @@
+# fwildclusterboot 0.14
+
+## Breaking Changes
+
+- the `print.boottest()` and `print.mboottest()` method have been deprecated, as both did not have a distinct use case. 
+- Bugfix: `boottest()` should never have run with `fixest::feols()` and 
+  varying slopes syntax via `var1[var2]`. Unfortunately it did for the heteroskedastic bootstrap - it's a bug. I am very sorry if you are affected by this! This version adds an error message for this case.
+
+## Performance
+
+Version 0.14 ...
+
+- sparsifies the "fast and reliable" bootstraps - bootstrap types 31, 33, 13 (which leads to good speed gains for problems with high dimensional fixed effects)
+- allows to project out cluster fixed effects when running the "fast and reliable" algorithms "11" and "31"
+- computes the generalized inverse `pinv` via rcpp eigen instead of `MASS::ginv()` whenever `Matrix::solve()` fails 
+- unlocks parallelization (nthreads was internally set to 1 for some reason)
+
+
+## rOpenSci Review feedback 
+
+- update docs:
+  - add a vignette on wild bootstrap concepts (wild bootstrap 101)
+  - better explanation of plot method in docs and vignette 
+  - some guidelines on how to turn messages and warnings off
+- reorganization of ropensci ssr tags into code
+- it is now possible to interrupt rcpp loops
+
+
+## Misc
+
+- throws a clear error message when the subcluster bootstrap is tried for the fast and reliable algos (currently not supported)
+- bumps the required `WildBootTests.jl` version to `0.9.7`
+
+
 # fwildclusterboot 0.13
 
 ## Potentially Breaking Changes: 
@@ -308,7 +342,7 @@ I have spent some time to clean up `fwildclusterboot's` internals, which should
 + Bug fix: for one-sided hypotheses for the WRU bootstrap (if impose_null = FALSE), the returned p-values were incorrect - they were reported as 'p', but should have been '1-p'. E.g. if the reported p-values was reported as 0.4, it should have been reported as 0.6. 
 + A new function argument `ssc` gives more control over the small sample adjustments made within `boottest()`. It closely mirrors the `ssc` argument in `fixest`. The only difference is that `fwildclusterboot::boot_ssc()'s` `fixef.K` argument currently has only one option, `'none'`, which means that the fixed effect parameters are discarded when calculating the number of estimated parameters k. 
 The default argument of `boot_ssc()` are `adj = TRUE, fixef.K = "none", cluster.adj = TRUE` and `cluster.df = "conventional"`. In fixest, the `cluster.df` argument is `"min"` by default. 
-Prior to v 0.6, by default, no small sample adjustments regarding the sample size N and the number of estimated parameters k were applied. The changes in v0.6 may slightly affect the output of `boottest()`. For exact reproducibility of previous results, set `adj = FALSE`. Setting `adj = TRUE` will not affect p-values and confidence intervals for *oneway clustering*, but the internally calculated t-stat, which is divided by $\sqrt{(N-k)/(N-1)}$. For *twoway* clustering, it might affect the number and order of invalid bootstrapped t-statistics (due to non-positive definite covariance matrices) and, through this channel, affect bootstrapped inferential parameters.
+Prior to v 0.6, by default, no small sample adjustments regarding the sample size N and the number of estimated parameters k were applied. The changes in v0.6 may slightly affect the output of `boottest()`. For exact reproducibility of previous results, set `adj = FALSE`. Setting `adj = TRUE` will not affect p-values and confidence intervals for *oneway clustering*, but the internally calculated t-stat, which is divided by sqrt(N-k)/(N-1). For *twoway* clustering, it might affect the number and order of invalid bootstrapped t-statistics (due to non-positive definite covariance matrices) and, through this channel, affect bootstrapped inferential parameters.
 
 + Testing: unit tests are now run on [github actions](https://github.com/s3alfisc/fwildclusterboot/actions/workflows/R-CMD-check.yaml) against [wildboottestjlr](https://github.com/s3alfisc/wildboottestjlr), which is a [JuliaConnectoR](https://github.com/stefan-m-lenz/JuliaConnectoR) based wrapper around [WildBootTests.jl](https://github.com/droodman/WildBootTests.jl), a Julia implementation of the fast wild cluster bootstrap algorithm. 
 + Additionally, minor speed tweaks. 

R/RcppExports.R

@@ -9,6 +9,24 @@ boot_algo3_crv3 <- function(B, G, k, v, scores_mat, scores_boot, inv_tXX_tXgXg,
     .Call('_fwildclusterboot_boot_algo3_crv3', PACKAGE = 'fwildclusterboot', B, G, k, v, scores_mat, scores_boot, inv_tXX_tXgXg, cores, R, delta_b_star)
 }
 
+convertSparse <- function(mat) {
+    .Call('_fwildclusterboot_convertSparse', PACKAGE = 'fwildclusterboot', mat)
+}
+
+compute_H <- function(G, R, tXXinv, tXgXg, scores_list, cores) {
+    .Call('_fwildclusterboot_compute_H', PACKAGE = 'fwildclusterboot', G, R, tXXinv, tXgXg, scores_list, cores)
+}
+
+pinv <- function(X) {
+    .Call('_fwildclusterboot_pinv', PACKAGE = 'fwildclusterboot', X)
+}
+
+#' Moore-Penrose Pseudo Inverses via Eigen
+#' @param A a matrix
+#' @return A matrix. Pseudo-Inverse of A.
+#' @noRd
+NULL
+
 #' Matrix Multiplication via Eigen
 #' @param A A matrix.
 #' @param B A matrix.
@@ -19,6 +37,10 @@ eigenMapMatMult <- function(A, B, nthreads) {
     .Call('_fwildclusterboot_eigenMapMatMult', PACKAGE = 'fwildclusterboot', A, B, nthreads)
 }
 
+eigen_pinv <- function(A) {
+    .Call('_fwildclusterboot_eigen_pinv', PACKAGE = 'fwildclusterboot', A)
+}
+
 #' Get maximum number of threads on hardware for open mp support
 #' @noRd
 cpp_get_nb_threads <- function() {