dataCleaning.Rmd

---
title: "Data Cleaning"
author: "Aleksander Brynjulf Hübert"
date: "04/05/2020"
output: html_document
---

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


```{r, warning=FALSE}
library(tidyverse)
library(ggplot2)
```

```{r}
bankData <- read.delim("data/bank-full.csv", sep =";" ,header = T, strip.white = TRUE)

#creates a dataset that keeps NAs as 'unknown'
bankDataUnkown <- bankData

#changes 'unknown' to NA
bankData[bankData == "unknown"] <- NA

sum(is.na(bankData))

sum(is.na(bankData$contact))
sum(is.na(bankData$poutcome))

#for testing the logistic regression without NAs
bankData <- bankData[,-16]

bankData <- na.omit(bankData)

ggplot(bankData, aes(y))+
    geom_bar()
```


```{r}
library(corrplot)
##correlations of NA values will go here

corrmat <-  model.matrix(~. -1, bankDataUnkown)

correlations <- cor(corrmat)

corrplot(correlations)
```

```{r}

```