Lab1RMD.Rmd

---
title: "Lab1RMD"
author: "Brandon Kim, Shreya Ravilla"
date: '2023-04-06'
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(warning = FALSE, message = FALSE) 
library(tidyverse)
```


```{r qqplot function in tidy}
qqcreatetidy <- function(x) {
  normalized_values <- sort(rnorm(length(x), mean = mean(x), sd = sd(x))) 
  # Create a sorted vector of random normally distributed values with the mean and sd of the inputted vector
  vector <- sort(x)
  df <- data.frame(vector, normalized_values)
  # since both vectors are sorted, having them put together in a dataframe will match the values from low to high.
  
  plt <- df %>%
    ggplot(aes(vector, normalized_values)) +
    geom_point() +
    geom_abline(intercept = 0, slope = 1) +
    ggtitle("qqplot")
  # Create a scatterplot with a line y = x. 
  
  return(plt)
}
```


```{r qqplot function in base}
qqcreatebase <- function(x) {
  normalized_values <- sort(rnorm(length(x), mean = mean(x), sd = sd(x))) 
  vector <- sort(x)
  
  plt <- plot(vector, normalized_values)
  abline(a = 0, b = 1)
  # add a y = x line
  title("qqplot")
  return(plt)
}
```


```{r readcsv}
df_oscars <- read_csv(here::here("Oscars-demographics-DFE.csv"))
head(df_oscars)
```


```{r warmup 1 tidy}
df_oscars %>% 
  filter(award %in% c("Best Director", 
                       "Best Actor", 
                       "Best Actress", 
                       "Best Supporting Actor", 
                       "Best Supporting Actress")) %>%
  group_by(movie) %>%
  summarise(unique = nlevels(factor(award))) %>%
  filter(unique == max(unique))
```


```{r warmup 1 base}
oscars_filtered <- df_oscars[df_oscars$award %in% c("Best Director", 
                                                    "Best Actor",
                                                    "Best Actress",
                                                    "Best Supporting Actor",
                                                    "Best Supporting Actress"),]

oscars_grouped <- aggregate(award ~ movie, data = oscars_filtered, FUN = function(x) length(unique(x)))

oscars_grouped[oscars_grouped$award == max(oscars_grouped$award), ]
```


```{r warmup 2 tidy}
df_oscars %>%
  filter(award == "Best Actress") %>%
  mutate(first_name = sub(" .*", "", person)) %>%
  count(first_name) %>%
  filter(n == max(n))
```


```{r warmup 2 base}

oscars_actress <- subset(df_oscars, award == "Best Actress")
oscars_actress$first_name <- sub(" .*", "", oscars_actress$person)
num_name <- table(oscars_actress$first_name)
max_name <- max(num_name)
subset(as.data.frame(num_name), Freq == max_name)
```

```{r warmup 3 tidy}
df_oscars %>% 
  mutate(birthplace = ifelse(birthplace=="New York City", "Ny", birthplace),
         region = sub(".*, ", "", birthplace)) %>%
  count(region) %>%
  filter(n == max(n))
```


```{r warmup 3 base}
df_oscars$birthplace <- ifelse(df_oscars$birthplace == "New York City", "Ny", df_oscars$birthplace)
df_oscars$region <- sub(".*, ", "", df_oscars$birthplace)
region_counts <- table(df_oscars$region)
max_region <- names(region_counts)[region_counts == max(region_counts)]
data.frame(region = max_region, n = max(region_counts))
```


```{r age and gender data cleaning}
library(lubridate)   
df_oscars <- df_oscars %>%
  mutate(award_date = as.Date(paste(as.character(year_of_award), "-02-01", sep = "")), 
         date_of_birth = as.Date(date_of_birth, "%d-%b-%Y"), 
         date_of_birth = update(date_of_birth, year = as.numeric(format(date_of_birth, "%Y")) + 
                                ifelse(as.numeric(format(date_of_birth, "%Y")) < 99, 1900, 0)),
         age = floor(time_length(difftime(award_date, date_of_birth), "years")), 
         years_since_inception = year_of_award - 1927)
```


```{r malelm}
df_males <- df_oscars %>%
  filter(award %in% c("Best Actor", "Best Supporting Actor")) 

lm_male <- lm(age ~ years_since_inception, data = df_males)
summary(lm_male)
```
For male actors, our linear model is:
$$
\hat{age}=43.47876+0.07471(year)
$$
where year is in years after the inception of the Oscars. With a p-value of 0.05551, we can conclude that there is moderate evidence to support that there is a relationship between years after inception and age for male award winners. This means that for every 1 year past inception, the predicted age of Best Actor and Best Supporting Actor winners will go up by 0.07471 years. 


```{r femalelm}
df_females <- df_oscars %>%
  filter(award %in% c("Best Actress", "Best Supporting Actress")) 

lm_female <- lm(age ~ years_since_inception, data = df_females)
summary(lm_female)
```

For female actors, our linear model is:
$$
\hat{age}=35.88847+0.03270(year)
$$
where year is in years after the inception of the Oscars. With a p-value of 0.4011, we can conclude that there is very weak evidence to suggest 
is a relationship between years after inception and age for female award winners. However, if a relationship did exist between typical age and years since inception, we can conclude that for every 1 year past inception, the predicted age of Best Actress and Best Supporting Actress winners will go up by 0.03270 years.