adds asserts to pd

 - checks all inputs
 - adds na checking for plotting functions
 - updated roxygen2

NAMESPACE

@@ -1,11 +1,16 @@
-# Generated by roxygen2 (4.0.1): do not edit by hand
+# Generated by roxygen2 (4.0.2): do not edit by hand
 
 export(ivar_points)
 export(partial_dependence)
 export(plot_imp)
 export(plot_twoway_partial)
 import(ggplot2)
 importFrom(assertthat,assert_that)
+importFrom(assertthat,is.count)
+importFrom(assertthat,is.flag)
+importFrom(assertthat,is.string)
+importFrom(assertthat,noNA)
+importFrom(assertthat,on_failure)
 importFrom(foreach,"%do%")
 importFrom(foreach,"%dopar%")
 importFrom(foreach,foreach)

R/pd.R

@@ -4,6 +4,7 @@
 #' from a fitted random forest object from the Party, randomForest, or randomForestSRC packages
 #'
 #' @importFrom foreach foreach %dopar% %do%
+#' @importFrom assertthat assert_that on_failure is.string is.count is.flag noNA
 #' 
 #' @param fit an object of class 'RandomForest-class' returned from \code{cforest}, an object
 #' of class 'randomForest' returned from \code{randomForest}, or an object of class 'rfsrc'
@@ -74,13 +75,20 @@
 #' }
 #' @export
 partial_dependence <- function(fit, df, var, cutoff = 10, empirical = TRUE) {
+    assert_that(any(class(fit) %in% c("RandomForest", "randomForest", "randomForestSRC")))
+    assert_that(is.data.frame(df))
+    assert_that(is.string(var))
+    assert_that(is.count(cutoff))
+    assert_that(is.flag(empirical))
+    assert_that(cutoff > nrow)
+
     if (any(class(fit) == "RandomForest")) {
         df <- data.frame(get("input", fit@data@env), get("response", fit@data@env))
         type <- class(df[, ncol(df)])
         y <- colnames(df[, ncol(df)])
         pkg <- "party"
-    }
-    else if (any(class(fit) == "randomForest")) {
+    } else if (any(class(fit) == "randomForest")) {
+        assert_that(noNA(df))
         type <- attr(fit$terms, "dataClasses")[1]
         y <- attr(attr(fit$terms, "dataClasses"), "names")[1]
         pkg <- "randomForest"
@@ -95,7 +103,7 @@ partial_dependence <- function(fit, df, var, cutoff = 10, empirical = TRUE) {
             type <- "survival"
         else
             type <- class(df[, y])
-    } else stop("Unsupported fit object class")
+    }
 
     rng <- expand.grid(lapply(var, function(x) ivar_points(df, x, cutoff, empirical)))
 

R/plot.R

@@ -1,7 +1,7 @@
-#' Plot permutation importance from random 
+#' Plot permutation importance from random forests
 #'
 #' @import ggplot2
-#' @importFrom assertthat assert_that
+#' @importFrom assertthat assert_that noNA
 #'
 #' @param var character or factor vector of variable labels
 #' @param imp numeric vector of variable permutation importance estimates
@@ -16,6 +16,8 @@ plot_imp <- function(var, imp, ylab = NULL, xlab = NULL, title = NULL) {
     assert_that(is.factor(var) | is.character(var))
     assert_that(length(var) == length(imp))
     assert_that(!(any(is.na(var) | any(is.na(imp)))))
+    assert_that(noNA(var))
+    assert_that(noNA(imp))
 
     df <- data.frame(imp, var)
     df$var <- factor(df$var, levels = df$var[order(df$imp)])
@@ -56,7 +58,12 @@ plot_twoway_partial <- function(var, pred, var_lab, grid, smooth = FALSE,
     assert_that(is.numeric(pred) | is.factor(pred) | is.integer(pred))
     assert_that(length(var) == length(pred))
     assert_that(!(any(is.na(var) | any(is.na(pred)))))
+    assert_that(noNA(var))
+    assert_that(noNA(pred))
+    assert_that(noNA(grid))
+    assert_that(is.flag(smooth))
     if (!missing(var_lab)) {
+        assert_that(noNA(var_lab))
         assert_that(is.factor(var_lab) | is.character(var_lab))
         assert_that(!missing(grid))
         assert_that(is.integer(grid) & length(grid == 2))

man/ivar_points.Rd

@@ -1,4 +1,4 @@
-% Generated by roxygen2 (4.0.1): do not edit by hand
+% Generated by roxygen2 (4.0.2): do not edit by hand
 \name{ivar_points}
 \alias{ivar_points}
 \title{Creates a prediction vector for variables to decrease computation time}

man/partial_dependence.Rd

@@ -1,4 +1,4 @@
-% Generated by roxygen2 (4.0.1): do not edit by hand
+% Generated by roxygen2 (4.0.2): do not edit by hand
 \name{partial_dependence}
 \alias{partial_dependence}
 \title{Partial dependence using random forests}

man/plot_imp.Rd

@@ -1,7 +1,7 @@
-% Generated by roxygen2 (4.0.1): do not edit by hand
+% Generated by roxygen2 (4.0.2): do not edit by hand
 \name{plot_imp}
 \alias{plot_imp}
-\title{Plot permutation importance from random}
+\title{Plot permutation importance from random forests}
 \usage{
 plot_imp(var, imp, ylab = NULL, xlab = NULL, title = NULL)
 }
@@ -20,6 +20,6 @@ plot_imp(var, imp, ylab = NULL, xlab = NULL, title = NULL)
 a ggplot2 object
 }
 \description{
-Plot permutation importance from random
+Plot permutation importance from random forests
 }
 

man/plot_twoway_partial.Rd

@@ -1,4 +1,4 @@
-% Generated by roxygen2 (4.0.1): do not edit by hand
+% Generated by roxygen2 (4.0.2): do not edit by hand
 \name{plot_twoway_partial}
 \alias{plot_twoway_partial}
 \title{Plot two-way partial dependence from random forests}