1-clean-data.Rmd

---
title: "Clean data"
output: html_document
---

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

## Import data

```{r import_data}
data_raw_dir = "data-raw/"
filename = "processed.cleveland.csv"
full_file = paste0(data_raw_dir, filename)

if (!file.exists(full_file)) {
  url = "http://archive.ics.uci.edu/ml/machine-learning-databases/heart-disease/processed.cleveland.data"
  # Load the remote data.
  data = rio::import(url, header = FALSE, format = "csv", na.strings = '?')
  
  # Update column names.
  names(data) = c("age", "sex", "cp", "trestbps", "chol","fbs", "restecg",
                "thalach","exang", "oldpeak","slope", "ca", "thal", "num")
  
  # Save to local directory.
  rio::export(data, file = full_file)
} else {
  data = rio::import(full_file)
}
```

## Categoricals to factors

```{r categoricals_to_factors}

# Explicitly define certain variables as factors rather than numbers.
data = ck37r::categoricals_to_factors(data,
              categoricals = c("ca", "cp", "slope", "thal"),
              verbose = TRUE)

str(data)
```

## Save unimputed version

```{r save_unimputed}
save(data, file = "data/clean-data-unimputed.RData")
```

## Missing values

```{r missing_values}

# We have a few missing values.
colSums(is.na(data))

# Omit missing values for now.
data = na.omit(data)

colSums(is.na(data))
```

## Save result

```{r save_data}
save(data, file = "data/clean-data-imputed.RData")
```