Review of the third day’s material

# Review of the third day’s material

---

---

# Reproducible documents

---

## Markdown syntax

# Header 1
## Header 2
### Header 3

Text **bold**, *italics*

- list 1
- list 2
```
]

Footnote[^1]

[^1]: Footnote content
![image caption](path/to/image.png)

[Link](https://google.com)

|   | Column 1 | Column 2 |
|:--|----:|--------:|
| Row 1 | Cell | Cell |
| Row 2 | Cell | Cell |
```
]

---

## R Markdown code chunks

````markdown
---
title: "My report"
author: "Me!"
bibliography: my_references.bib
output:
    html_document:
        theme: sandstone
---

Cite: [@Hoejsgaard2006a]

Code chunk.

```{r chunk-label, chunk.options}
# R Code here
plot(iris)
library(knitr)
# Table
kable(iris)
```
````

---

# Efficient coding

---

## read_csv repeats twice, wrangling four times

```r
*nhanes_2009 <- read_csv(here::here("data/nhanes-2009_10.csv"))
nhanes_2009 %>% 
*   mutate(ProblemBMIs = !between(BMI, 18.5, 40)) %>%
*   filter(!is.na(ProblemBMIs)) %>%
*   select(Age, Poverty, Pulse, BPSysAve, BPDiaAve, TotChol,
*          SleepHrsNight, PhysActiveDays, ProblemBMIs) %>%
*   gather(Measurement, Value, -ProblemBMIs) %>%
*   na.omit() %>%
    ggplot(aes(y = Value, x = Measurement, colour = ProblemBMIs)) +
    geom_jitter(position = position_dodge(width = 0.6)) +
    scale_color_viridis_d(end = 0.8) +
    labs(y = "", x = "") +
    theme_minimal() +
    theme(legend.position = c(0.85, 0.85)) +
    coord_flip()
```

---

## jitter repeats twice

```r
nhanes_2009 <- read_csv(here::here("data/nhanes-2009_10.csv")) 
nhanes_2009 %>% 
    mutate(ProblemBMIs = !between(BMI, 18.5, 40)) %>% 
    filter(!is.na(ProblemBMIs)) %>% 
    select(Age, Poverty, Pulse, BPSysAve, BPDiaAve, TotChol,
           SleepHrsNight, PhysActiveDays, ProblemBMIs) %>% 
    gather(Measurement, Value, -ProblemBMIs) %>% 
    na.omit() %>% 
*   ggplot(aes(y = Value, x = Measurement, colour = ProblemBMIs)) +
*   geom_jitter(position = position_dodge(width = 0.6)) +
*   scale_color_viridis_d(end = 0.8) +
*   labs(y = "", x = "") +
*   theme_minimal() +
*   theme(legend.position = c(0.85, 0.85)) +
*   coord_flip()
```

---

## Move code into functions

```r
*read_mutate_gather <- function(.file_path) {
*   .file_path %>%
        read_csv() %>% 
        mutate(ProblemBMIs = !between(BMI, 18.5, 40)) %>% 
        filter(!is.na(ProblemBMIs)) %>% 
        select(Age, Poverty, Pulse, BPSysAve, BPDiaAve, TotChol,
               SleepHrsNight, PhysActiveDays, ProblemBMIs) %>% 
        gather(Measurement, Value, -ProblemBMIs) %>% 
        na.omit() 
*} 
*plot_jitter <- function(.dataset) {
*   .dataset %>%
        ggplot(aes(y = Value, x = Measurement, colour = ProblemBMIs)) +
        geom_jitter(position = position_dodge(width = 0.6)) +
        scale_color_viridis_d(end = 0.8) +
        labs(y = "", x = "") +
        theme_minimal() +
        theme(legend.position = c(0.85, 0.85)) +
        coord_flip()
*} 
```

---

## density repeats twice

```r
nhanes_2009 %>% 
    mutate(ProblemBMIs = !between(BMI, 18.5, 40)) %>% 
    filter(!is.na(ProblemBMIs)) %>% 
    select(Age, Poverty, Pulse, BPSysAve, BPDiaAve, TotChol,
           SleepHrsNight, PhysActiveDays, ProblemBMIs) %>% 
    gather(Measurement, Value, -ProblemBMIs) %>% 
    na.omit() %>% 
*   ggplot(aes(x = Value, fill = ProblemBMIs)) +
*   geom_density(alpha = 0.35) +
*   facet_wrap(~Measurement, scales = "free") +
*   scale_fill_viridis_d(end = 0.8) +
*   labs(y = "", x = "") +
*   theme_minimal() +
*   theme(legend.position = c(0.85, 0.15),
*         strip.text = element_text(face = "bold"))
```

---

## Move code into functions

```r
*plot_density <- function(.dataset) {
*   .dataset %>%
        ggplot(aes(x = Value, fill = ProblemBMIs)) +
        geom_density(alpha = 0.35) +
        facet_wrap(~Measurement, scales = "free") +
        scale_fill_viridis_d(end = 0.8) +
        labs(y = "", x = "") +
        theme_minimal() +
        theme(legend.position = c(0.85, 0.15),
              strip.text = element_text(face = "bold")) 
*} 
```

---

## Two dataframes, two figures each.

```r
# Start with file paths:
files <- 
    c(here::here("data/nhanes-2009_10.csv"),
      here::here("data/nhanes-2011_12.csv"))
```

```r
# Apply wrangling to each data file:
data_list <- 
    files %>% 
    map(read_mutate_gather)
```

```r
# Apply figure to each data file:
# Plot the jitters
map(data_list, plot_jitter)

# Plot the density
map(data_list, plot_density)
```

---

## Parallel processing

```r
# Start with file paths:
files <- 
    c(here::here("data/nhanes-2009_10.csv"),
      here::here("data/nhanes-2011_12.csv"))
*library(furrr)
*plan(multiprocess)
# Apply wrangling to each data file:
data_list <- 
    files %>% 
*   future_map(read_mutate_gather)
```

```r
# Apply figure to each data file:
# Plot the jitters
*future_map(data_list, plot_jitter)

# Plot the density
*future_map(data_list, plot_density)
```