EDA1.Rmd

---
title: "Exploratory Data Analysis"
author: "OUYANG YUCHEN"
output: html_document
---

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

## Install & Import Libraries
In order to implement EDA, two libraries are needed. **readxl** is for reading the excel file and the very important package **ggplot2** under **tidyverse** is used for data visulization.

```{r initialization}
# install.packages("readxl","tidyverse")
library(readxl)
library(tidyverse)
```

```{r import-dataset}
df <- read_excel("Features1D/Features/Features.xlsx")
df
```

## Scale and Key

```{r scale}
ggplot(df,aes(x=scale,fill=scale)) + 
  geom_bar() + 
  facet_wrap(~mood) +
  labs(title = "Scale vs. Mood")
```
```{r key}
ggplot(df,aes(x=key,fill=key)) + 
  geom_bar() + 
  facet_wrap(~mood) +
  labs(title = "Key vs. Mood")
```

## Tempo

```{r Tempo}
ggplot(df,aes(y=tempo,x=mood,fill=mood)) + 
  geom_boxplot() + 
  labs(title = "Tempo vs. Mood")
```
```{r Tempo-hist}
ggplot(df,aes(x=tempo)) + 
  geom_histogram(binwidth = 6, color = "skyblue3", fill = "skyblue") + 
  geom_density(aes(y=after_stat(count)* 6),color = "skyblue4") + 
  facet_wrap(~mood) + 
  labs(title = "Tempo vs. Mood", y = "")
```

## RMS

```{r RMS-avg}
ggplot(df,aes(x = mood, y=rms_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "RMS vs. Mood", y = "average RMS")
```

```{r RMS-avg-hist}
ggplot(df,aes(x=rms_mean)) + 
  geom_histogram(binwidth = 0.02, color="skyblue3",fill = "skyblue") + 
  geom_density(aes(y=after_stat(count)* 0.02),color="skyblue4")+
  facet_wrap(~mood) +
  labs(title = "RMS vs. Mood", x = "average RMS", y = "") 
```

## Tonnetz

```{r tonnetz}
ggplot(df,aes(x = mood, y=tonnetz_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "Tonnetz vs. Mood", y = "Tonnetz Mean")
```

```{r RMS-var-hist}
ggplot(df,aes( x=tonnetz_mean, group = mood)) + 
  geom_histogram(fill="skyblue", color = "skyblue3", binwidth = 0.01) +
  geom_density(aes(y=..count..* 0.01),color = "skyblue4")+
  facet_wrap(~mood)+
  labs(title = "Tonnetz vs. Mood", x = "Average Tonnetz", y = "")
```

## Centroid

```{r cent}
ggplot(df,aes(x = mood, y=centroid_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "Centroid vs. Mood", y = "Centroid Mean")
```

```{r cent-hist}
ggplot(df,aes(x=centroid_mean)) + 
  geom_histogram(fill="skyblue", color = "skyblue3", binwidth = 100) +
  geom_density(aes(y=..count..* 100),color = "skyblue4")+
  facet_wrap(~mood)+
  labs(title = "Centroid vs. Mood", x = "Centroid Mean", y="")
```

## Spectral Roll-off

```{r roll-off}
ggplot(df,aes(x = mood, y=rolloff_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "Roll-off vs. Mood", y = "Roll-off Mean")
```

```{r rolloff-hist}
ggplot(df,aes(x=chroma_mean)) + 
  geom_histogram(fill="skyblue", color = "skyblue3", binwidth = 0.02) +
  geom_density(aes(y=..count..*0.02),color = "skyblue4")+
  facet_wrap(~mood)+
  labs(title = "Roll-off vs. Mood", x = "Roll-off Mean", y = "")
```

## Chroma

```{r chroma}
ggplot(df,aes(x = mood, y=chroma_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "Chroma vs. Mood", y = "Chroma Mean")
```

```{r chroma-hist}
ggplot(df,aes(x=chroma_mean)) + 
  geom_histogram(fill="skyblue", color = "skyblue3", binwidth = 0.02) +
  geom_density(aes(y=..count..* 0.02),color = "skyblue4")+
  facet_wrap(~mood)+
  labs(title = "Chroma vs. Mood", x = "Chorma Mean", y="")
```

## Zero Crossing Rate

```{r zcr}
ggplot(df,aes(x = mood, y=zcr_mean, fill = mood)) + 
  geom_boxplot() + 
  labs(title = "Zcr vs. Mood", y = "Zcr Mean")
```

```{r zcr-hist}
ggplot(df,aes(x=zcr_mean)) + 
  geom_histogram(fill="skyblue", color = "skyblue3", binwidth = 0.01) +
  geom_density(aes(y=..count..* 0.01),color = "skyblue4")+
  facet_wrap(~mood)+
  labs(title = "Zcr vs. Mood", x = "Zcr Mean", y="")
```