num -> numeric, prob -> classprob for #65

Author: topepo

NAMESPACE

@@ -19,8 +19,8 @@ S3method(predict_classprob,"_lognet")
 S3method(predict_classprob,"_multnet")
 S3method(predict_classprob,model_fit)
 S3method(predict_confint,model_fit)
-S3method(predict_num,"_elnet")
-S3method(predict_num,model_fit)
+S3method(predict_numeric,"_elnet")
+S3method(predict_numeric,model_fit)
 S3method(predict_predint,model_fit)
 S3method(predict_quantile,model_fit)
 S3method(predict_raw,"_elnet")
@@ -92,8 +92,8 @@ export(predict_classprob)
 export(predict_classprob.model_fit)
 export(predict_confint)
 export(predict_confint.model_fit)
-export(predict_num)
-export(predict_num.model_fit)
+export(predict_numeric)
+export(predict_numeric.model_fit)
 export(predict_predint)
 export(predict_predint.model_fit)
 export(predict_quantile)

R/boost_tree.R

@@ -394,7 +394,7 @@ xgb_by_tree <- function(tree, object, new_data, type, ...) {
       pred <- boost_tree_xgboost_data$classes$post(pred, object)
       pred <- tibble(.pred = factor(pred, levels = object$lvl))
     } else {
-      pred <- boost_tree_xgboost_data$prob$post(pred, object)
+      pred <- boost_tree_xgboost_data$classprob$post(pred, object)
       pred <- as_tibble(pred)
       names(pred) <- paste0(".pred_", names(pred))
     }

R/boost_tree_data.R

@@ -31,7 +31,7 @@ boost_tree_xgboost_data <-
           verbose = 0
         )
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = NULL,
       func = c(fun = "xgb_pred"),
@@ -58,7 +58,7 @@ boost_tree_xgboost_data <-
           newdata = quote(new_data)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         if (is.vector(x)) {
@@ -106,7 +106,7 @@ boost_tree_C5.0_data <-
         newdata = quote(new_data)
       )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         as_tibble(x)
@@ -142,7 +142,7 @@ boost_tree_spark_data <-
           seed = expr(sample.int(10^5, 1))
         )
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = format_spark_num,
       func = c(pkg = "sparklyr", fun = "ml_predict"),
@@ -162,7 +162,7 @@ boost_tree_spark_data <-
           dataset = quote(new_data)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = format_spark_probs,
       func = c(pkg = "sparklyr", fun = "ml_predict"),

R/linear_reg.R

@@ -219,12 +219,12 @@ predict._elnet <-
   }
 
 #' @export
-predict_num._elnet <- function(object, new_data, ...) {
+predict_numeric._elnet <- function(object, new_data, ...) {
   if (any(names(enquos(...)) == "newdata"))
     stop("Did you mean to use `new_data` instead of `newdata`?", call. = FALSE)
 
   object$spec <- eval_args(object$spec)
-  predict_num.model_fit(object, new_data = new_data, ...)
+  predict_numeric.model_fit(object, new_data = new_data, ...)
 }
 
 #' @export

R/linear_reg_data.R

@@ -30,7 +30,7 @@ linear_reg_lm_data <-
       func = c(pkg = "stats", fun = "lm"),
       defaults = list()
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = NULL,
       func = c(fun = "predict"),
@@ -100,7 +100,7 @@ linear_reg_glmnet_data <-
           family = "gaussian"
         )
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = organize_glmnet_pred,
       func = c(fun = "predict"),
@@ -135,7 +135,7 @@ linear_reg_stan_data <-
           family = expr(stats::gaussian)
         )
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = NULL,
       func = c(fun = "predict"),
@@ -224,7 +224,7 @@ linear_reg_spark_data <-
       protect = c("x", "formula", "weight_col"),
       func = c(pkg = "sparklyr", fun = "ml_linear_regression")
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = function(results, object) {
         results <- dplyr::rename(results, pred = prediction)

R/logistic_reg_data.R

@@ -44,7 +44,7 @@ logistic_reg_glm_data <-
           type = "response"
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         x <- tibble(v1 = 1 - x, v2 = x)
@@ -121,7 +121,7 @@ logistic_reg_glmnet_data <-
           s = quote(object$spec$args$penalty)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = organize_glmnet_prob,
       func = c(fun = "predict"),
@@ -170,7 +170,7 @@ logistic_reg_stan_data <-
           newdata = quote(new_data)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         x <- object$fit$family$linkinv(x)
@@ -278,7 +278,7 @@ logistic_reg_spark_data <-
           dataset = quote(new_data)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = format_spark_probs,
       func = c(pkg = "sparklyr", fun = "ml_predict"),

R/mars_data.R

@@ -23,7 +23,7 @@ mars_earth_data <-
       func = c(pkg = "earth", fun = "earth"),
       defaults = list(keepxy = TRUE)
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = maybe_multivariate,
       func = c(fun = "predict"),
@@ -48,7 +48,7 @@ mars_earth_data <-
           type = "response"
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         x <- x[,1]

R/mlp_data.R

@@ -25,7 +25,7 @@ mlp_keras_data <-
       func = c(pkg = "parsnip", fun = "keras_mlp"),
       defaults = list()
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = maybe_multivariate,
       func = c(fun = "predict"),
@@ -47,7 +47,7 @@ mlp_keras_data <-
           x = quote(as.matrix(new_data))
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = function(x, object) {
         x <- as_tibble(x)
@@ -81,7 +81,7 @@ mlp_nnet_data <-
       func = c(pkg = "nnet", fun = "nnet"),
       defaults = list(trace = FALSE)
     ),
-    pred = list(
+    numeric = list(
       pre = NULL,
       post = maybe_multivariate,
       func = c(fun = "predict"),

R/multinom_reg_data.R

@@ -40,7 +40,7 @@ multinom_reg_glmnet_data <-
           s = quote(object$spec$args$penalty)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = check_glmnet_lambda,
       post = organize_multnet_prob,
       func = c(fun = "predict"),
@@ -85,7 +85,7 @@ multinom_reg_spark_data <-
           dataset = quote(new_data)
         )
     ),
-    prob = list(
+    classprob = list(
       pre = NULL,
       post = format_spark_probs,
       func = c(pkg = "sparklyr", fun = "ml_predict"),

R/nearest_neighbor_data.R

@@ -22,8 +22,8 @@ nearest_neighbor_kknn_data <-
       func = c(pkg = "kknn", fun = "train.kknn"),
       defaults = list()
     ),
-    pred = list(
-      # seems unnecessary here as the predict_num catches it based on the
+    numeric = list(
+      # seems unnecessary here as the predict_numeric catches it based on the
       # model mode
       pre = function(x, object) {
         if (object$fit$response != "continuous") {
@@ -60,7 +60,7 @@ nearest_neighbor_kknn_data <-
           type = "raw"
         )
     ),
-    prob = list(
+    classprob = list(
       pre = function(x, object) {
         if (!(object$fit$response %in% c("ordinal", "nominal"))) {
           stop("`kknn` model does not appear to use class predictions. Was ",