---
title: "Spotify 2018"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    vertical_layout: fill
    social: [ "twitter", "facebook", "menu"]
    source_code: embed
---

```{r setup, include=FALSE}
library(flexdashboard)
library(knitr)
library(DT)
library(rpivotTable)
library(openintro)
library(highcharter)
library(ggvis)

library(pacman)
pacman::p_load(pacman,dplyr, ggplot2, plotly,tidyr, rio, gridExtra, scales, ggcorrplot, caret)
```


```{r}
df <- import("./top2018.csv")
df$hit <- with(df, paste(artists,",", name))
df$key <- as.factor(df$key)
df$mode <- as.factor(df$mode)
df$time_signature <- as.factor(df$time_signature)
```

```{r}
mycolors <- c("blue", "#FFC125", "darkgreen", "darkorange")
```

Key Indicators
=====================================

Row
-------------------------------------

### Best Spotify Artist 2018

```{r}
valueBox(df$artists[3],
         icon = "fa-user")
```

### **Number of Hits in Top 100**

```{r}
gauge(round(6,
            digits = 2),
            min = 0,
            max = 100,
            gaugeSectors(success = c(5, 100),
                         warning = c(2, 5),
                         danger = c(0, 2),
                         colors = c("green", "yellow", "red")))
```


Row
-------------------------------

### Sound Keys Counts (Pitch Class notation)

```{r}
c <- sum(df$key==0)
c2 <- sum(df$key==1)
d <- sum(df$key==2)
d2 <- sum(df$key==3)
e <- sum(df$key==4)
f <- sum(df$key==5)
f2 <- sum(df$key==6)
g <- sum(df$key==7)
g2 <- sum(df$key==8)
a <- sum(df$key==9)
a2 <- sum(df$key==10)
b <- sum(df$key==11)

scale = c(c,c2,d,d2,e,f,f2,g,g2,a,a2,b)

plot_ly(
  x = c('C','C#','D','D#','E','F','F#','G','G#','A','A#','B'),
  y = scale,
  type = "bar"
)
```

### Trak Modality

```{r}
m1 <- sum(df$mode==0)
m2 <- sum(df$mode==1)

plot_ly(
  x = c('Minor','Major'),
  y = c(m1,m2),
  type = "bar"
)
```

### Time Signature (how many beats are in each bar)

```{r}
t1 <- sum(df$time_signature==3)
t2 <- sum(df$time_signature==4)
t3 <- sum(df$time_signature==5)

plot_ly(
  x = c('3','4','5'),
  y = c(t1,t2,t3),
  type = "bar"
)
```

Overall Analysis
========================================

### Artists with most hits in Top 100

```{r}
df %>%
  group_by(artists) %>%
  summarise(freq = n()) %>%
  arrange(desc(freq)) %>%
  slice(1:10) %>%
  ggplot(., aes(reorder(artists, +freq), freq))+
  geom_bar(stat = "identity", fill = "royalblue1", col = "grey10")+
  coord_flip()+
  labs(x = "" ,y = "Top 10 2018 Artists")+
  geom_text(aes(label = freq, y = freq/2))
```

### Features Correlation

```{r}
df$key <- as.numeric(df$key)
df$mode <- as.numeric(df$mode)
df$time_signature <- as.numeric(df$time_signature)
corr <- round(cor(df[,4:16]), 8)
ggcorrplot(corr)
```

Music Analysis
========================================

### Danceability

```{r}
df %>%
  arrange(desc(danceability)) %>%
  slice(1:5) %>%
  ggplot(., aes(reorder(hit, +danceability), danceability))+
  geom_bar(stat = "identity", fill = "royalblue1", col = "grey10")+
  coord_flip()+
  labs(x = "" ,y = "Score")+
  geom_text(aes(label = danceability, y = danceability/2))
```

### Energy

```{r}
df %>%
  arrange(desc(energy)) %>%
  slice(1:5) %>%
  ggplot(., aes(reorder(hit, + energy), energy))+
  geom_bar(stat = "identity", fill = "royalblue1", col = "grey10")+
  coord_flip()+
  labs(x = "" ,y = "Score")+
  geom_text(aes(label = energy, y = energy/2))
```

### Loudness

```{r}
df %>%
  arrange(desc(loudness)) %>%
  slice(1:5) %>%
  ggplot(., aes(reorder(hit, + loudness), loudness))+
  geom_bar(stat = "identity", fill = "royalblue1", col = "grey10")+
  coord_flip()+
  labs(x = "" ,y = "Score")+
  geom_text(aes(label = loudness, y = loudness/2))
```

Time Analysis
========================================

### Duration (mins:secs)

```{r}
df %>%
  arrange(desc(duration_ms)) %>%
  slice(1:5) %>%
  ggplot(., aes(reorder(hit, + duration_ms), duration_ms))+
  geom_bar(stat = "identity", fill = "royalblue1", col = "grey10")+
  coord_flip()+
  labs(x = "" ,y = "Score")+
  geom_text(aes(label = format(as.POSIXct(Sys.Date())+duration_ms/1000, "%M:%S"), y = duration_ms/2))
```

### Musical Tempo Bands 

```{r}
q <- sum(df$tempo > 168)
w <- sum(df$tempo >= 110 & df$tempo <= 168)
e <- sum(df$tempo >= 76 & df$tempo <= 108)
r <- sum(df$tempo >= 66 & df$tempo <= 76)
t <- sum(df$tempo < 65)

plot_ly(
  x = c('Presto', 'Allegro', 'Andante','Adagio', 'Lenght'),
  y = c(q,w,e,r,t),
  type = "bar"
)
```

Dataset Table
========================================

```{r}
datatable(df[,2:16],
          caption = "Spotify Top 100 2018",
          rownames = T,
          filter = "top",
          options = list(pageLength = 25))
```

Pivot Table
=========================================

```{r}
rpivotTable(df,
            aggregatorName = "Count",
            cols= "fm",
            rows = "State",
            rendererName = "Heatmap")
```

Summary 
===========================================

This report provides a summary of the [Top Spotify Tracks of 2018 Kaggle Dataset](https://www.kaggle.com/nadintamer/top-spotify-tracks-of-2018). 


This dashboard was realised using R Shiny Markdown. 


* Click [here](https://ppiconsulting.dev/work.html) to come back my Personal Website.

* Click [here](https://ppiconsulting.dev/blog/) to visit my Personal Blog.

* Click [here](https://github.com/pierpaolo28/R-Programming) to access this Dashboard Code Repository.