3.workflows-solutions.Rmd

---
title: "Solutions to exercise: workflows, filtering and sorting"
author: "Mark Dunning and Matt Eldridge"
date: '`r format(Sys.time(), "Last modified: %d %b %Y")`'
output: html_document
---

## Part I -- Workflows using pipes

1. Read in the patients dataset and rewrite the following cleaning steps as a workflow using the `%>%` operator.

```{r eval = FALSE}
library(tidyverse)
patients <- read_tsv("patient-data.txt")
patients <- mutate(patients, Smokes = Smokes %in% c("TRUE", "Yes"))
patients <- mutate(patients, Height = as.numeric(str_remove(Height, pattern = "cm$")))
patients <- mutate(patients, Weight = as.numeric(str_remove(Weight, pattern = "kg$")))
patients <- mutate(patients, BMI = Weight / (Height / 100) ** 2)
patients <- mutate(patients, Overweight = BMI > 25)
```

```{r}
library(tidyverse)
patients <- read_tsv("patient-data.txt") %>%
  mutate(Smokes = Smokes %in% c("TRUE", "Yes")) %>%
  mutate(Height = as.numeric(str_remove(Height, pattern = "cm$"))) %>%
  mutate(Weight = as.numeric(str_remove(Weight, pattern = "kg$"))) %>%
  mutate(BMI = Weight / (Height / 100) ** 2) %>%
  mutate(Overweight = BMI > 25)
patients
```

2. Add a step to the workflow to round the Height, Weight and BMI to 1 decimal place.

```{r}
patients <- read_tsv("patient-data.txt") %>%
  mutate(Smokes = Smokes %in% c("TRUE", "Yes")) %>%
  mutate(Height = as.numeric(str_remove(Height, pattern = "cm$"))) %>%
  mutate(Weight = as.numeric(str_remove(Weight, pattern = "kg$"))) %>%
  mutate(BMI = Weight / (Height / 100) ** 2) %>%
  mutate(Overweight = BMI > 25) %>%
  mutate_at(vars(Height, Weight, BMI), round, digits = 1)
patients
```

## Part II - Filtering rows

3. Filter for female patients from New York or New Jersey.

```{r}
filter(patients, Sex == "Female", State == "New York" | State == "New Jersey")
filter(patients, Sex == "Female", State %in% c("New York", "New Jersey"))
filter(patients, Sex == "Female", str_starts(State, "New "))
```

4. Filter for overweight smokers that are still alive.

```{r}
filter(patients, Overweight & Smokes == "Smoker" & !Died)
```