---
title: "LAB 2: Perception et couleurs"
author: "Antoine Neuraz"
date: "19/11/2019"
output: 
  xaringan::moon_reader:
    css: ['default','css/my_style.css']
    lib_dir: libs
    seal: false
    nature:
      ratio: '4:3'
      countIncrementalSlides: false
      self-contained: true
      beforeInit: "addons/macros.js"
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, fig.asp = .6)
library(ggplot2)
#library(showtext)
library(vizoR)
library(dplyr)
library(patchwork)
library(see) 
library(RColorBrewer)
```

class: center, middle, title

# Lab 2: Perception et couleurs

### 2019-2020

## Dr. Antoine Neuraz

### AHU Informatique médicale
#### Hôpital Necker-Enfants malades, </br> Université de Paris


---
class: inverse, center, middle
# Perception des différentes marques dans ggplot2

---
## TODO: échauffement

#### Générer un dataset aléatoire avec la fonction vizoR::generate_dataset_uniform
```{r}
size <- list(100, 2)
min_x <- 0
max_x <- 1
seed <- 34
```
---
## TODO: perception

#### 1. Réaliser des plots avec les échelles suivantes sur la variable group : 
- couleur
- forme
- angle
- taille
- luminosité
- courbe
- encapsulage
- remplissage

Certaines échelles sont très simples à mettre en place (e.g. couleur, forme) mais d'autres n'existent pas directement. Il faut trouver une alternative. 

#### 2. Comparer l'efficacité des différentes échelles pour distinguer les 2 groupes

---
## Couleur


```{r, echo = F}

dt <- generate_dataset_uniform(
  dataset_size = size,
  min_x = min_x,
  max_x = max_x,
  seed = seed
)
```

--
.small[
```{r, echo = T}

p_color <- ggplot( data = dt, 
                   aes(x = x, y = y, color = group)) +
  geom_point(size = 3, alpha = .6) +
  see::scale_color_material_d() +
  vizoR::theme_void_complete() +
  labs(subtitle= "Couleur")

p_color
```
]


---
## Angle 

--

.small[
```{r}
p_angle <- dt %>%
  mutate(angle = ifelse(group == "group1", 0, pi / 3)) %>%
  ggplot(
    data = ., aes( x = x, y = y, angle = angle ) ) +
  geom_spoke(radius = 0.02,size = .8, alpha = .6 ) +
  theme_void_complete() +
  scale_color_material_d() +
  ggtitle("Angle")
p_angle
```
]
---
## Taille

--

.small[
```{r}
p_size <- dt %>%
  mutate(size = ifelse(group == "group1", 2, 3)) %>%
  ggplot( data = .,
          aes( x = x,y = y,size = size)) +
  geom_point(alpha = .6) +
  theme_void_complete() +
  scale_size(range = c(1, 3)) +
  ggtitle("Taille")
p_size
```
]

---
## Luminosité

--

.small[
```{r}
p_grey <- dt %>%
  ggplot(
    data = .,aes(x = x,y = y,color = group)) +
  geom_point(size = 3, alpha = .6) +
  theme_void_complete() +
  #scale_color_grey() +
  scale_color_grey(start=.8, end=.2)+
  ggtitle("Luminosité")
p_grey
```
]
---
## Courbe

--

.small[
```{r}
dt <- dt %>%
  mutate(curvature = ifelse(group == "group1", 0, 1))

p_curve <- dt %>%
  ggplot(data = .,aes(x = x,y = y,xend = x, yend = y+max_x/50, curvature = curvature)) +
  #geom_curve()+
  geom_curve(data = subset(dt, group == 'group1'), curvature = 0, alpha = .7) +
  geom_curve(data = subset(dt, group == 'group2'), curvature = .7, alpha = .7) +
  scale_color_material_d() +
  theme_void_complete() +
  ggtitle("Courbe")
p_curve
```
]
---
## Encapsulage

--

.small[
```{r}
p_box <- dt %>%
  ggplot(data = .,aes(x = x,xend = x+max_x/50,y = y,yend = y, group = group)) +
  geom_point(data = subset(dt, group=='group2'),aes(x = x+max_x/100), shape = 22, size = 13) + 
  geom_segment() +
  scale_color_material_d() +
  theme_void_complete() +
  ggtitle("Encapsulage")
p_box
```
]

---
## Forme

--

.small[
```{r}
p_shape <- dt %>%
  ggplot(data = .,aes(x = x,y = y,shape = group)) +
  geom_point(size = 3, alpha = .6) +
  theme_void_complete() +
  ggtitle("Forme")
p_shape
```
]
---
## Remplissage

--

.small[
```{r}
p_fill <- dt %>%
  ggplot(data = ., aes( x = x,y = y,fill = group)) +
  geom_point(size = 3, shape = 21, alpha = .7) +
  scale_fill_manual(values = c('group2' = 'black', 'group1' = 'white')) +
  theme_void_complete() +
  ggtitle("Remplissage")
p_fill
```
]

---
class: full, center, middle
```{r, echo = F,  out.width = '2000px'}
p_color + p_angle + 
  p_size + p_grey + 
  p_curve + p_box + 
  p_shape + p_fill + 
  plot_layout(ncol = 2) 


```


---
class: full
### Les couleurs dans ggplot2 `display.brewer.all()`

![:scale 80%](img/palettes.jpg)

---
## TODO: couleurs

#### Charger le dataset diamonds et créer un sous-dataset aléatoire de 1000 lignes

#### Plot carat en fonction du prix et de la couleur

#### changer la palette par défaut vers une autre palette disponible

---
```{r}
dsamp <- diamonds[sample(nrow(diamonds), 1000), ]
ggplot(dsamp, aes(carat, price)) +
  geom_point(aes(colour = color)) +
  scale_color_brewer(palette = "Set3") + 
  facet_wrap(~color)
```


---
## TODO: couleurs 2

#### Plot carat en fonction du prix avec carat en double encodage

#### Aller sur [http://colorbrewer2.org]() et trouver une palette divergente

#### Créer une palette custom basée sur cette palette et l'appliquer au plot précédent

#### Caler la palette sur le carat moyen 

#### Annoter le plot avec une ligne désignant le carat moyen et un texte expliquant cette ligne
---


```{r}

ggplot(dsamp, aes(carat, price)) +
  geom_point(aes(colour = carat)) +
  scale_color_distiller(palette="RdYlBu")
```

---
```{r, eval = F}

#showtext_auto()
#font_add_google("Schoolbell", "bell")

font_family = "sans"
annotate_color = "grey50"

midpoint = (max(dsamp$carat)-min(dsamp$carat))/2

ggplot(dsamp, aes(carat, price)) +
  geom_vline(xintercept = midpoint, color = annotate_color) +
  geom_point(aes(colour = carat)) +
  scale_color_gradient2(low = "#d8b365", 
                        mid="#f5f5f5", 
                        high="#5ab4ac", 
                        midpoint = midpoint) +
  annotate("text", 
           x=.78, y=15000, hjust=1, srt=40,
           label ="this is the midpoint",  
           family=font_family, 
           color=annotate_color) +
  annotate("curve", 
           x = .8, xend=midpoint-.01, y=15000, yend = 14000,
           curvature = -.5, 
           color=annotate_color ,
           arrow=arrow(length = unit(0.03, "npc") )) + 

  theme_elegant() +
  theme(panel.grid.minor = element_blank(), 
        panel.grid.major.x = element_blank(), 
        legend.position = "none") 
```

---
```{r, echo = F}

#showtext_auto()
#font_add_google("Schoolbell", "bell")

font_family = "sans"
annotate_color = "grey50"

midpoint = (max(dsamp$carat)-min(dsamp$carat))/2

ggplot(dsamp, aes(carat, price)) +
  geom_vline(xintercept = midpoint, color = annotate_color) +
  geom_point(aes(colour = carat)) +
  scale_color_gradient2(low = "#d8b365", 
                        mid="#f5f5f5", 
                        high="#5ab4ac", 
                        midpoint = midpoint) +
  annotate("text", 
           x=.78, y=15000, hjust=1, srt=40,
           label ="this is the midpoint",  
           family=font_family, 
           color=annotate_color) +
  annotate("curve", 
           x = .8, xend=midpoint-.01, y=15000, yend = 14000,
           curvature = -.5, 
           color=annotate_color ,
           arrow=arrow(length = unit(0.03, "npc") )) + 

  theme_elegant() +
  theme(panel.grid.minor = element_blank(), 
        panel.grid.major.x = element_blank(), 
        legend.position = "none") 
```