candy_data_analysis.Rmd

---
title: "candy_data_analysis"
author: "VM"
date: "11/11/2020"
output: 
  html_document:
    number_sections: yes
    toc: yes
    toc_float: yes
---

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

# Load libraries and read the file

```{r}
library(tidyverse)
full_candy_data <- read_csv("~/dirty_data_project/task4/clean_data/full_clean_candy_data.csv")
```
# Q1
## What is the total number of candy ratings given across the three years. (number of candy ratings, not number of raters. Don’t count missing values)

```{r}
full_candy_data %>% 
  select(year, comments)  %>% 
  drop_na() %>% 
  nrow()
```
```{r}
full_candy_data %>% 
  select(year, comments)  %>% 
  filter(comments == c("despair", "joy", "meh")) %>% 
  nrow()
```
# Q2
## What was the average age of people who are going out trick or treating and the average age of people 3. not going trick or treating?
```{r}
full_candy_data %>% 
  select(age, going_or_not) %>% 
  filter(going_or_not == "yes") %>% 
  summarise(avg_age_going = mean(age)) %>% 
  round()

full_candy_data %>% 
  select(age, going_or_not) %>% 
  filter(going_or_not == "no") %>% 
  summarise(avg_age_not_going = mean(age)) %>% 
  round()

```
# Q3
## For each of joy, despair and meh, which candy bar revived the most of these ratings?
```{r}
frequency_candies <- full_candy_data %>% 
  select(candies, comments) %>% 
  filter(comments == c("despair", "joy", "meh")) %>% 
  group_by(comments, candies) %>% 
  count()

frequency_candies %>% 
  group_by(comments) %>% 
  slice_max(order_by = n)

```
# Q4
## How many people rated Starburst as despair?
```{r}
full_candy_data %>% 
  filter(candies == "starburst") %>% 
  filter(comments == "despair") %>% 
  nrow()
```
# Q5
## For the next three questions, count despair as -1, joy as +1 and meh as 0.
```{r}
rating_data <- full_candy_data  %>% 
  filter(comments == c("despair", "joy", "meh")) %>% 
  mutate(rating_count = case_when(
    comments == "despair" ~ -1,
    comments == "joy" ~ +1,
    comments == "meh" ~ 0
  ))

rating_data
```

## What was the most popular candy bar by this rating system for each gender in the dataset?
```{r}
rating_data %>% 
  select(gender, candies, rating_count) %>% 
  filter(gender == c("male", "female")) %>% 
  group_by(gender,candies) %>% 
  summarise(total = sum(rating_count)) %>% 
  slice_max(total)
  
```

## What was the most popular candy bar in each year?
```{r}
rating_data %>% 
  select(year, rating_count, candies) %>% 
  group_by(year, candies) %>% 
  summarise(total = sum(rating_count)) %>% 
  slice_max(total)
```

## What was the most popular candy bar by this rating for people in US, Canada, UK and all other countries?

```{r}
rating_data %>% 
  select(country, candies, rating_count) %>% 
  drop_na(country) %>%
  group_by(country, candies) %>% 
  summarise(total = sum(rating_count)) %>% 
  slice_max(order_by = total, with_ties = FALSE)

```
```{r}
rating_data %>% 
  select(country, candies, rating_count) %>% 
  mutate(country = case_when(
    country == "uk" ~ "uk",
    country == "usa" ~ "usa",
    country == "canada" ~ "canada",
    TRUE ~ "other" 
  )) %>% 
  group_by(country, candies) %>% 
  summarise(total = sum(rating_count)) %>% 
  slice_max(total)

```