Full_Data.Rmd

---
title: "Full Data"
author: "Jeremiah Sagers"
date: "2024-04-25"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

library(tidyverse)
library(mice)
```

```{r}
# Read in Data
house <- read.csv("train.csv")
response <- house$SalePrice # stored for later
testdata <- read.csv("test.csv")
full <- rbind(house[-81], testdata) # combine train and test

NAtoNone <- c(2,3,7,10,24,25,26,31,32,33,34,36,54,56,58,59,61,64,65,73,74,75,79)
for (i in NAtoNone) {
  fulltmp <- full
  fulltmp[i][is.na(fulltmp[i])] <- "None"
  full <- fulltmp
}

full <- full[-1380,]
response <- response[-1380]

full <- full %>% mutate_if(is.character, as.factor)
full <- full %>% mutate_at(c('MSSubClass','MoSold', 'OverallQual', 'OverallCond'), as_factor)


# I don't add YrMoSold because we don't need it for the models, plus my model performed better without it anyway (on validation, at least).  If you do add YrMoSold, the imputation and separation still works.
#YrMoSold <- do.call(paste, c(sep='',list(full$YrSold),list(rep("-", times = length(full$YrSold))),
#                             list(full$MoSold),list(rep("-01", times = length(full$YrSold)))))
#YrMoSold <- as_date(YrMoSold)
#full <- as.data.frame(append(full, list(YrMoSold = YrMoSold), after = 78))

full <- full[-c(1,60)] # remove GarageYrBlt

# Separate the data back into train and test and add back the response variable
house <- full[1:1459,]
testdata <- full[1460:2918,]
house$SalePrice <- response

# Imputation must be done after separation due to the randomness of the values imputed.  If done to "full", there will be differences between the resulting train dataset and the one given by ReadData.R.
numhouse <- select_if(house, is.numeric)
imputed_house <- mice(numhouse,
                      m = 10,
                      method = 'pmm',
                      seed = 1)

set.seed(1)
imphouse <- complete(imputed_house)
housetmp <- house
s = 1
for (i in names(imphouse)[1:length(names(imphouse))]) { # use 1:length because "Id" has already been removed
  housetmp[i] <- imphouse[i]
}
house <- housetmp

numtest <- select_if(testdata, is.numeric)
imputed_test <- mice(numtest,
                      m = 10,
                      method = 'pmm',
                      seed = 1)

set.seed(1)
imptest <- complete(imputed_test)
testtmp <- testdata
s = 1
for (i in names(imptest)[1:length(names(imptest))]) {
  testtmp[i] <- imptest[i]
}
testdata <- testtmp

md.pattern(house)
md.pattern(testdata)
```