module4.Rmd

---
output:
  pdf_document: default
  html_document: default
---
start your engines (load your packages)
```{r}
library(tidyverse)

theme_set(theme_classic(base_size = 20)) 
```

read in the files
```{r}
movies_imdb=read_delim("movies/movies_imdb.txt",delim=",")
movies_rottentom=read_delim("movies/movies_rottentom.txt",delim=",")
```

read in drinks and convert it to long format with just beer, wine, and spirits
```{r}
drinks=read_delim("drinks/drinks_US.tsv",delim="\t")

drinks_long=drinks %>% 
  select(-total) %>%
  pivot_longer(names_to="type",values_to="gallons",cols=beer:spirits)

drinks_long

ggplot(data=drinks_long,aes(x=year,y=gallons))+geom_point()
```

make some plots (what can we do to make this better?)
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons))+geom_point()
```

line plot? wow -that's weird....
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons))+geom_line()
```

what if we use the "group" mapping... hmmm 
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,group=type))+geom_line()
```

let's add some color so we know what drink is what
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,group=type,col=type))+geom_line()

## note that when you use "color" or "fill" it automatically passes that same argument to "group", so this line is effectively the same
ggplot(data=drinks_long,aes(x=year,y=gallons,col=type))+geom_line()
```

What about better colors? (What's wrong with this code? Why doesn't it work?)
```{r}
#ggplot(data=drinks_long,aes(x=year,y=gallons,col=type))+geom_line()+scale_color_manual(values = c("black","red"))
```

What about better colors?
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,col=type))+geom_line()+scale_color_brewer(palette = "Set1")
```

What about conveying more information? (why are the colors not the RColorBrewer palette colors?)
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_color_brewer(palette = "Set1")
```

well that didn't work as well as we thought:
BOOM:
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_brewer(palette = "Set1")
```

we can also change the hue and saturation manually
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_hue(l=45,c=50)

ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_hue(l=45)

```

##Faceting
```{r}
ggplot(data=drinks_long,aes(x=year,y=gallons))+geom_line()+facet_grid(~type)
```

arranging multiple figures with gridArrange
```{r}
library(gridExtra)
plot1=ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_hue(l=45)
plot2=ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_hue(l=45,c=50)
grid.arrange(plot1,plot2,nrow=1)
```

Can we save these figures? How should we save them?
```{r}
drinks_plot=ggplot(data=drinks_long,aes(x=year,y=gallons,fill=type))+geom_area()+scale_fill_brewer(palette = "Set1")
drinks_plot
ggsave(drinks_plot,file="~/Downloads/drinks.pdf",width=10,height = 4)
```

##STRINGS ON STRINGS ON STRINGS

We're going to use the stringr package

using str_detect and str_subset: which one returns a logical?
```{r}
str_detect(movies_imdb$genres,"Action")
str_subset(movies_imdb$genres,"Action")
str_subset(movies_imdb$genres,"action")  ## what's up with this output? Why is it different than the line above? 
```

stringr and filter()
```{r}
action_movies = movies_imdb %>% filter(str_detect(genres,"Action"))
action_movies

## looks like more action movies have been made recently, but is this true? What's our null expectation given the data? 
ggplot(action_movies,aes(x=title_year))+geom_histogram() + xlab("movie release (year)")
```

What about movies with a lead actor whose name begins with "Ch"
```{r}
movies_imdb %>% filter(str_detect(actor_1_name,"^Ch"))%>% select(actor_1_name,title_year,movie_title,everything())
```

What about movies with a lead actor whose name ends with "son"
```{r}
movies_imdb %>% filter(str_detect(actor_1_name,"on$"))%>% select(actor_1_name,title_year,movie_title,everything())
```

EVERYTHING IS CASE SENSITIVE
use str_to_lower() and str_to_upper() to fix that


Dates are fun
```{r}
library(lubridate)
ymd(20101215)
ymd("2010/12/15")
ymd("2010Dec15")
ymd("2010December15")
date1=ymd("2010 Dec 15")

year(date1)
month(date1)
mday(date1)
```