Chapters_24_to_25.Rmd

---
title: "Chapters 24 and 25"
author: "Laura"
date: "12/18/2019"
output: html_document
---

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

library(tidyverse); library(skimr); library(nycflights13); library(GGally); library(ggstance); library(lvplot); library(hexbin); library(modelr); library(magrittr); library(maps); library(stringr); library(htmlwidgets); library(forcats); library(mapproj); library(lubridate); library(babynames); library(splines)

```

# Notes for Chapters 24 and 25

## Chapter 24: Model Building

```{r}
flights %>% 
  mutate(date = make_date(year, month, day)) %>% 
  count(date) %>% 
  ggplot(aes(date, n)) +
    geom_line()

daily <- flights %>% 
  mutate(date = make_date(year, month, day)) %>% 
  count(date)

daily %>% 
  mutate(wday = wday(date, label = TRUE)) %>% 
  ggplot(aes(wday, n)) +
    geom_boxplot()

weekly <- daily %>% 
  mutate(wday = wday(date, label = TRUE))

mod <- lm(n ~ wday, data = weekly)

# predictions
grid <- weekly %>% 
  data_grid(wday) %>% 
  add_predictions(mod, "n")
grid

weekly %>% 
  ggplot(aes(wday, n)) +
    geom_boxplot() +
    geom_point(data = grid, color = "red", size = 4)

# residuals
res <- weekly %>% 
  add_residuals(mod, "resid")
  
res %>% 
  ggplot(aes(date, resid)) +
    geom_line(aes(color = wday)) +
    geom_ref_line(h = 0, colour = "black", size = 0.5) +
    scale_colour_viridis_d() +
    theme_bw() +
    labs(color = "Weekday", title = "Bla")
 
res %>% 
  ggplot(aes(date, resid)) +
    geom_smooth(se = FALSE, span = 0.1) +
    geom_line(color = "grey") +
    geom_ref_line(h = 0, colour = "black", size = 0.5)
   

```

```{r}
weekly %>% 
  filter(wday == "Sat") %>% 
  ggplot(aes(date, n)) +
    geom_line() +
    geom_point() +
    scale_x_date(date_breaks = "1 month", date_labels = "%b")

terms <- function(date){
  cut(date,
      breaks = ymd(20130101, 20130605, 20130825, 20140101),
      labels = c("spring", "summer", "fall"))
}

terms <- weekly %>% 
  mutate(term = terms(date)) 

terms%>% 
  filter(wday == "Sat") %>% 
  ggplot(aes(date, n, color = term)) +
    geom_line() +
    geom_point() +
    scale_x_date(date_breaks = "1 month", date_labels = "%b")

terms$wday <-  fct_relevel(terms$wday, "Sun", after = 7)

terms %>% 
  ggplot(aes(wday, n, color = term)) +
    geom_boxplot()
  
```
```{r}
terms

mod1 <- lm(n ~ wday, data = terms)
mod2 <- lm(n ~ wday * term, data = terms)

terms %>% 
  gather_residuals(wterm = mod1, woterm = mod2) %>% 
  ggplot(aes(date, resid, color = model)) +
    geom_line()

grid <- terms %>% 
  data_grid(wday, term) %>% 
  add_predictions(mod1, "n")

terms %>% 
  ggplot(aes(wday, n)) +
    geom_boxplot() +
    geom_point(data = grid, color = "red") +
    facet_grid(cols = vars(term)) +
    theme(axis.text.x = element_text(angle = 90, hjust = 1))

mod3 <- MASS::rlm(n ~ wday * term, data = terms)

rterms <- terms %>% 
  add_residuals(mod3) 

rterms %>% ggplot(aes(date, resid)) +
    geom_line()
  
  
```
```{r}

mod <- MASS::rlm(n ~ wday * ns(date, 5), data = weekly)

weekly %>% 
  data_grid(wday, date = seq_range(date, by = 13)) %>% 
  add_predictions(mod, "pred") %>% 
  ggplot(aes(date, pred, color = wday)) +
    geom_line() +
    geom_point() 

mod1 <- MASS::rlm(n ~ wday * date, data = weekly)

weekly %>% 
  data_grid(wday, date = seq_range(date, by = 13)) %>% 
  add_predictions(mod1, "pred") %>% 
  ggplot(aes(date, pred, color = wday)) +
    geom_line() +
    geom_point() 
```

## Chapter 19: Functions