don't attempt to impute covariates if there is no missingness

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

@@ -29,7 +29,7 @@ jobs:
 
     steps:
       - name: Checkout repo
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
 
       - name: Setup R
         uses: r-lib/actions/setup-r@v2
@@ -50,7 +50,7 @@ jobs:
 
       - name: Install package dependencies
         run: |
-          install.packages(c("remotes", "devtools", "rcmdcheck", "covr", "sessioninfo"))
+          install.packages(c("remotes", "rcmdcheck", "covr", "sessioninfo"))
           remotes::install_deps(dependencies = TRUE)
         shell: Rscript {0}
 

DESCRIPTION

@@ -33,7 +33,7 @@ Depends: R (>= 3.6.0)
 Imports:
     data.table,
     assertthat,
-    origami (>= 1.0.3),
+    origami (>= 1.0.7),
     R6,
     uuid,
     BBmisc,
@@ -68,7 +68,7 @@ Suggests:
     glmnet,
     grf,
     gbm,
-    hal9001 (>= 0.4.4),
+    hal9001 (>= 0.4.6),
     h2o,
     keras,
     nloptr,
@@ -84,16 +84,14 @@ Suggests:
     xgboost,
     lightgbm,
     dbarts,
-    gam (>= 1.15.0),
-    haldensify (>= 0.2.3),
+    gam,
+    haldensify (>= 0.2.7),
     mgcv,
     hts,
     GA,
     SIS,
     partykit
 Remotes:
-    github::tlverse/origami,
-    github::tlverse/hal9001@devel,
     github::nhejazi/haldensify
 License: GPL-3
 Language: en-US

R/process_data.R

@@ -132,13 +132,15 @@ process_data <- function(data, nodes, column_names, flag = TRUE,
       # add inidicators to column map and covariate list
       column_names[missing_indicator_vars] <- missing_indicator_cols
       nodes$covariates <- c(nodes$covariates, missing_indicator_vars)
+
+      # impute covariates
+      data_missing_covars <- data[, missing_covar_cols, with = FALSE]
+      imputed <- impute(data.frame(data_missing_covars))
+
+      # update data
+      set(data, , missing_covar_cols, imputed)
+
     }
-    # impute covariates
-    data_missing_covars <- data[, missing_covar_cols, with = FALSE]
-    imputed <- impute(data.frame(data_missing_covars))
-
-    # update data
-    set(data, , missing_covar_cols, imputed)
   }
 
   na_Y <- (!is.null(nodes$outcome) && any(is.na(data[, outcome_columns, with = F])))

testresults.txt

@@ -0,0 +1,111 @@
+[2024-04-29 20:20:53] ══ Documenting ═════════════════════════════════════════════════════════════════
+[2024-04-29 20:20:53] ℹ Updating sl3 documentation
+[2024-04-29 20:20:54] ℹ Loading sl3
+[2024-04-29 20:20:57] 
+[2024-04-29 20:20:57] ══ Building ════════════════════════════════════════════════════════════════════
+[2024-04-29 20:20:57] Setting env vars:
+[2024-04-29 20:20:57] • CFLAGS    : -Wall -pedantic
+[2024-04-29 20:20:57] • CXXFLAGS  : -Wall -pedantic
+[2024-04-29 20:20:57] • CXX11FLAGS: -Wall -pedantic
+[2024-04-29 20:20:57] • CXX14FLAGS: -Wall -pedantic
+[2024-04-29 20:20:57] • CXX17FLAGS: -Wall -pedantic
+[2024-04-29 20:20:57] • CXX20FLAGS: -Wall -pedantic
+[2024-04-29 20:20:57] ── R CMD build ─────────────────────────────────────────────────────────────────
+[2024-04-29 20:20:57] * checking for file ‘/home/jrcoyle/Dropbox/tlverse/sl3/DESCRIPTION’ ... OK
+[2024-04-29 20:20:57] * preparing ‘sl3’:
+[2024-04-29 20:20:57] * checking DESCRIPTION meta-information ... OK
+[2024-04-29 20:20:57] * installing the package to process help pages
+[2024-04-29 20:21:14] Loading required namespace: sl3
+[2024-04-29 20:21:15] * saving partial Rd database
+[2024-04-29 20:21:19] * creating vignettes ... OK
+[2024-04-29 20:21:19] * checking for LF line-endings in source and make files and shell scripts
+[2024-04-29 20:21:19] * checking for empty or unneeded directories
+[2024-04-29 20:21:19] Removed empty directory ‘sl3/tests/testthat/_snaps’
+[2024-04-29 20:21:19] * building ‘sl3_1.4.5.tar.gz’
+[2024-04-29 20:21:19] 
+[2024-04-29 20:21:19] ══ Checking ════════════════════════════════════════════════════════════════════
+[2024-04-29 20:21:19] Setting env vars:
+[2024-04-29 20:21:19] • _R_CHECK_CRAN_INCOMING_USE_ASPELL_           : TRUE
+[2024-04-29 20:21:19] • _R_CHECK_CRAN_INCOMING_REMOTE_               : FALSE
+[2024-04-29 20:21:19] • _R_CHECK_CRAN_INCOMING_                      : FALSE
+[2024-04-29 20:21:19] • _R_CHECK_FORCE_SUGGESTS_                     : FALSE
+[2024-04-29 20:21:19] • _R_CHECK_PACKAGES_USED_IGNORE_UNUSED_IMPORTS_: FALSE
+[2024-04-29 20:21:19] • NOT_CRAN                                     : false
+[2024-04-29 20:21:19] ── R CMD check ─────────────────────────────────────────────────────────────────
+[2024-04-29 20:21:19] * using log directory ‘/tmp/RtmpZW35EK/file1237e20c4dd84/sl3.Rcheck’
+[2024-04-29 20:21:19] * using R version 4.3.3 (2024-02-29)
+[2024-04-29 20:21:19] * using platform: x86_64-pc-linux-gnu (64-bit)
+[2024-04-29 20:21:19] * R was compiled by
+[2024-04-29 20:21:19]     gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+[2024-04-29 20:21:19]     GNU Fortran (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+[2024-04-29 20:21:19] * running under: Pop!_OS 20.04 LTS
+[2024-04-29 20:21:19] * using session charset: UTF-8
+[2024-04-29 20:21:20] * using options ‘--no-manual --as-cran’
+[2024-04-29 20:21:20] * checking for file ‘sl3/DESCRIPTION’ ... OK
+[2024-04-29 20:21:20] * this is package ‘sl3’ version ‘1.4.5’
+[2024-04-29 20:21:20] * package encoding: UTF-8
+[2024-04-29 20:21:20] * checking package namespace information ... OK
+[2024-04-29 20:21:20] * checking package dependencies ... OK
+[2024-04-29 20:21:20] * checking if this is a source package ... OK
+[2024-04-29 20:21:20] * checking if there is a namespace ... OK
+[2024-04-29 20:21:21] * checking for executable files ... OK
+[2024-04-29 20:21:21] * checking for hidden files and directories ... OK
+[2024-04-29 20:21:21] * checking for portable file names ... OK
+[2024-04-29 20:21:21] * checking for sufficient/correct file permissions ... OK
+[2024-04-29 20:21:37] * checking whether package ‘sl3’ can be installed ... [16s/16s] OK
+[2024-04-29 20:21:37] * checking installed package size ... OK
+[2024-04-29 20:21:37] * checking package directory ... OK
+[2024-04-29 20:21:37] * checking for future file timestamps ... OK
+[2024-04-29 20:21:37] * checking ‘build’ directory ... OK
+[2024-04-29 20:21:37] * checking DESCRIPTION meta-information ... OK
+[2024-04-29 20:21:37] * checking top-level files ... OK
+[2024-04-29 20:21:37] * checking for left-over files ... OK
+[2024-04-29 20:21:37] * checking index information ... OK
+[2024-04-29 20:21:38] * checking package subdirectories ... OK
+[2024-04-29 20:21:38] * checking R files for non-ASCII characters ... OK
+[2024-04-29 20:21:38] * checking R files for syntax errors ... OK
+[2024-04-29 20:21:39] * checking whether the package can be loaded ... OK
+[2024-04-29 20:21:40] * checking whether the package can be loaded with stated dependencies ... OK
+[2024-04-29 20:21:41] * checking whether the package can be unloaded cleanly ... OK
+[2024-04-29 20:21:42] * checking whether the namespace can be loaded with stated dependencies ... OK
+[2024-04-29 20:21:44] * checking whether the namespace can be unloaded cleanly ... OK
+[2024-04-29 20:21:45] * checking loading without being on the library search path ... OK
+[2024-04-29 20:21:48] * checking startup messages can be suppressed ... OK
+[2024-04-29 20:21:49] * checking dependencies in R code ... OK
+[2024-04-29 20:21:50] * checking S3 generic/method consistency ... OK
+[2024-04-29 20:21:52] * checking replacement functions ... OK
+[2024-04-29 20:21:53] * checking foreign function calls ... OK
+[2024-04-29 20:22:00] * checking R code for possible problems ... OK
+[2024-04-29 20:22:02] * checking Rd files ... OK
+[2024-04-29 20:22:02] * checking Rd metadata ... OK
+[2024-04-29 20:22:02] * checking Rd line widths ... OK
+[2024-04-29 20:22:03] * checking Rd cross-references ... OK
+[2024-04-29 20:22:04] * checking for missing documentation entries ... OK
+[2024-04-29 20:22:08] * checking for code/documentation mismatches ... OK
+[2024-04-29 20:22:09] * checking Rd \usage sections ... OK
+[2024-04-29 20:22:09] * checking Rd contents ... OK
+[2024-04-29 20:22:10] * checking for unstated dependencies in examples ... OK
+[2024-04-29 20:22:10] * checking contents of ‘data’ directory ... OK
+[2024-04-29 20:22:10] * checking data for non-ASCII characters ... OK
+[2024-04-29 20:22:10] * checking LazyData ... OK
+[2024-04-29 20:22:10] * checking data for ASCII and uncompressed saves ... OK
+[2024-04-29 20:22:10] * checking installed files from ‘inst/doc’ ... OK
+[2024-04-29 20:22:10] * checking files in ‘vignettes’ ... OK
+[2024-04-29 20:22:41] * checking examples ... [30s/31s] OK
+[2024-04-29 20:22:44] * checking for unstated dependencies in ‘tests’ ... OK
+[2024-04-29 20:22:44] * checking tests ...
+[2024-04-29 20:28:05]   Running ‘testthat.R’ [401s/321s]
+[2024-04-29 20:28:05]  [401s/321s] OK
+[2024-04-29 20:28:05] * checking for unstated dependencies in vignettes ... OK
+[2024-04-29 20:28:05] * checking package vignettes in ‘inst/doc’ ... OK
+[2024-04-29 20:28:10] * checking re-building of vignette outputs ... OK
+[2024-04-29 20:28:10] * checking for non-standard things in the check directory ... OK
+[2024-04-29 20:28:10] * checking for detritus in the temp directory ... OK
+[2024-04-29 20:28:10] * DONE
+[2024-04-29 20:28:10] 
+[2024-04-29 20:28:10] Status: OK
+[2024-04-29 20:28:10] 
+[2024-04-29 20:28:11] ── R CMD check results ────────────────────────────────────────── sl3 1.4.5 ────
+[2024-04-29 20:28:11] Duration: 6m 51.5s
+[2024-04-29 20:28:11] 
+[2024-04-29 20:28:11] 0 errors ✔ | 0 warnings ✔ | 0 notes ✔

tests/testthat/test-missing_outcome.R

@@ -0,0 +1,32 @@
+context("Incorporating corrections for missingness in covariates")
+
+library(data.table)
+library(assertthat)
+library(uuid)
+library(sl3)
+library(tmle3)
+set.seed(34831)
+
+# setup data for test
+data(cpp)
+data <- as.data.table(cpp)
+data[, parity01 := as.numeric(data$parity > 0)]
+data[, parity01_fac := factor(data$parity01)]
+data[, haz01 := as.numeric(data$haz > 0)]
+
+node_list <- list(
+  W = c(
+    "apgar1", "apgar5", "gagebrth", "mage",
+    "meducyrs", "sexn"
+  ),
+  A = "waz",
+  Y = "haz01"
+)
+
+# drop missing A for now, might add back to test later
+missing_W <- apply(is.na(data[, c(node_list$W, node_list$A),
+  with = FALSE
+]), 1, any)
+data <- data[!missing_W]
+
+task <- sl3_Task$new(data, outcome=node_list$Y, covariates = node_list$W)