预处理函数 R答案

【问题标题】：preProcess function R预处理函数 R
【发布时间】：2021-05-09 23:07:30
【问题描述】：

我对 R 很陌生。

我正在尝试遵循以下程序：

imputedData <- preProcess( select(train, -SalePrice),
  method = c("center", "scale", "knnImpute", "nzv", 'YeoJohnson')
  )
#install.packages('RANN')
library(RANN)
trainTrans <- predict(imputedData, train)

我有这个错误

Must subset rows with a valid subscript vector. x Subscript nn$nn.idx must be a simple vector, not a matrix.

我已经安装了 Caret 包

火车数据集是我从 CSV 文件中导入的表格

【问题讨论】：

我无法重现您的错误。你真的在 predict() 遇到错误？

标签： r preprocessor

【解决方案1】：

这是程序

加载包

knitr::opts_chunk$set(echo = TRUE, cache = TRUE, message = FALSE, warning = FALSE)

library(tidyverse)
library(caret)
library(GGally)
library(lattice)
library(corrplot)
library(factoextra)
library(FactoMineR)
library(magrittr)
theme_set(theme_bw())
set.seed(181019)

数据加载

train <- readr::read_csv("train.csv")
test  <- readr::read_csv("test.csv")

缺失值

missing_threshold <- .4
is_too_scarce <- lapply(select(train, -SalePrice), function(x) mean(is.na(x)) > missing_threshold)
is_too_scarce <- map_lgl(select(train, -SalePrice), ~mean(is.na(.x)) > missing_threshold)
not_too_scarce <- names(is_too_scarce)[!is_too_scarce]
train <- select(train, SalePrice, not_too_scarce)
train %<>% select(SalePrice, not_too_scarce)
test  %<>% select(not_too_scarce)

预处理

imputedData <- preProcess( select(train, -SalePrice),
  method = c("center", "scale", "knnImpute", "nzv", 'YeoJohnson')
  )
#install.packages('RANN')
library(RANN)
testTrans  <- predict(imputedData, test)
trainTrans <- predict(imputedData, train)

【讨论】：