cake_data_analysis.Rmd

---
title: "cake_data_analysis"
author: "VM"
date: "09/11/2020"
output: 
  html_document:
    number_sections: yes
    toc: yes
    toc_float: yes
---

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

# Read data first and load libraries

```{r}
library(tidyverse)
library(here)
library(janitor)
clean_data <- read_csv("~/dirty_data_project/task2/clean_data/clean_cake_data.csv")
wide_data <- read_csv("~/dirty_data_project/task2/clean_data/clean_wide_cake_data.csv")
long_data <- read_csv("~/dirty_data_project/task2/clean_data/clean_long_cake_data.csv")
                      
```

```{r}
head(clean_data)
head(wide_data)
head(long_data)
```
There are a lot of NA as these ingredient were not used in cake, therefore, 
they were left and further ignored in analysis.

# Q1
## Which cake has the most cocoa in it?

```{r}
clean_data %>% 
  filter(ingredient == "cocoa") %>% 
  remove_empty("cols") 
  
wide_data %>% 
  select(cakes, cocoa_tablespoon) %>% 
  drop_na(cocoa_tablespoon) %>% 
  arrange(desc(cocoa_tablespoon)) %>% 
  head(1)

long_data %>% 
  group_by(ingredient) %>% 
  filter(ingredient == "cocoa") %>% 
  slice_max(value)
```
# Q2
## For sponge cake, how many cups of ingredients are used in total?
```{r}
clean_data %>% 
  select(measure, ingredient, sponge) %>% 
  filter(measure == "cup") %>% 
  drop_na() %>% 
  count()

wide_data %>% 
  filter(cakes == "sponge") %>% 
  remove_empty("cols") %>% 
  select(matches("cup$")) %>% 
  ncol()

long_data %>% 
  filter(cakes == "sponge") %>% 
  filter(measure == "cup") %>% 
  summarise(count = n_distinct(value))
```
# Q3
## How many ingredients are measured in teaspoons?
```{r}
clean_data %>% 
  filter(measure == "teaspoon") %>% 
  n_distinct()

wide_data %>% 
  select(matches("teaspoon")) %>% 
  ncol()

long_data %>% 
  group_by(ingredient) %>% 
  filter(measure == "teaspoon") %>% 
  summarise(count = n_distinct(ingredient)) %>% 
  count()
```
# Q4
## Which cake has the most unique ingredients?
```{r}
long_data %>% 
  group_by(cakes) %>% 
  drop_na() %>% 
  summarise(count = n_distinct(ingredient)) %>% 
  slice_max(count)
```

# Q5 
## Which ingredients are used only once?

```{r}
long_data %>% 
  group_by(ingredient) %>% 
  drop_na() %>% 
  count(ingredient) %>% 
  filter(n == 1)
```