DanielS_purrr_tidyproject.Rmd

---
title: "DATA607_tidy_purrr"
author: "Daniel Sullivan"
date: "4/11/2021"
output: html_document
---

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

## PURRR by beginers for beginers

The goal of this introduction is to review some of the different shortcuts that are implements within Purrr using the mtcars data.  


start by loading tidyverse which contains the purr package
```{r load libraries and data, message=FALSE, warning=FALSE}
library(tidyverse)
```
Then load in the mt cars data
```{r}
my_data <- mtcars
```



Purrr provides some very interesting additions to the apply functions. The modify and modify_if functions provide very easy to use functions that would be harder and longer to code with regular apply functions. additionally purrr provides shortcuts. This is seen below with the "~.+1" essentially I am using shortcuts to add one to each mpg total. with the tilda denoting the function to apply. 
```{r}
modify(my_data$mpg, ~.+1)
```

without the shortcuts the above lines would look like this which is much longer to right. 

```{r}
modify(my_data$mpg, function(x){1+x})
```
These shortcuts have multiple applications and can be used to do caulculations across rows as well. for example if i wanted to find mpg by weight.
```{r}
pmap(list(my_data$mpg, my_data$wt), ~..1/..2)
```
however purrr shines when you start to have to apply functions to lists of lists for example if we nest mtcars on whether the car has a v shaped engine or a straight engine we can use map functions where apply functions wouldnt work. 
```{r}
nest_data<-my_data%>%
  group_by(vs)%>%
  nest()
```
Thefirst line prints the data of cars that dont have a v shaped engine

the second prints the data frame for those that do.
```{r}
print(nest_data$data[1])
print(nest_data$data[2])
```

From there we can apply map functions to each nested data frame and get summary statistics for each. 
```{r}
Nest_mpg<- function(df){mean(df$mpg, na.rm = TRUE)}
Nest_wt<- function(df){median(df$wt, na.rm = TRUE)}

nest_data$avg_mpg<- nest_data$data%>%
  map(Nest_mpg)

nest_data$median_wt<- nest_data$data%>%
  map(Nest_wt)

```

There are many other built in functions that are usfull in purrr that can be used to apply to lists of lists or lists of data frames. 


**Additional examples, by Matt Lucich**

If you have ever run into issues working with nested data, purrr's pluck function may be something that saves you time in the future. It is intuitive in its ability to accept integer positions (column, row), but also allows you to specify string names and accessor functions. Note, that if pluck receives positions/strings that do not exist it will return null, to receive errors instead, be sure to use the chuck function.

```{r}

# View dataframe
nest_data

# View nested dataframe
nest_data$avg_mpg

# Select the third column
nest_data %>% pluck(3)

# Select the second column and first row (which returns the nested dataframe)
nest_data %>% pluck(2, 1)

# Select the second column and second row, then the "disp" column of the nested dataframe
nest_data %>% pluck(2, 2, "disp")

```