add lab7

5 年之前 · 27aa1e5527
--- a/courses/07_time_text.Rmd
+++ b/courses/07_time_text.Rmd
@@ -20,6 +20,16 @@ library(dplyr)
 library(stringi)
 ```

 class: center, middle, title

 # Visualisation de données textuelles et temporelles

 ### 2019-2020

 ## Dr. Antoine Neuraz

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

 ---
 class: center, full
--- a/courses/07_time_text.html
+++ b/courses/07_time_text.html
@@ -1,8 +1,8 @@
 <!DOCTYPE html>
 <html xmlns="http://www.w3.org/1999/xhtml" lang="" xml:lang="">
 <html>
  <head>
    <title>Cours 07 - Données temporelles et textuelles</title>
    <meta charset="utf-8" />
    <meta charset="utf-8">
    <meta name="author" content="Antoine Neuraz" />
    <link href="libs/remark-css-0.0.1/default.css" rel="stylesheet" />
    <link rel="stylesheet" href="css/my_style.css" type="text/css" />
@@ -162,11 +162,29 @@ class: full, center
 [interactive](https://visual.ly/community/interactive-graphic/business/house-hunting)

 ---
 class: center
 ## Gapminder 

 &lt;iframe width="800" height="400" src="https://www.youtube.com/embed/BPt8ElTQMIg" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen&gt;&lt;/iframe&gt;

 ---
 ## divide and conquer

 ![](img/walmart.png)

 ---
 ## Streamgraph

 ![](img/streamgraph.png)

 ---
 class: center
 ## Eventflow

 &lt;iframe width="800" height="400" src="https://www.youtube.com/embed/ZN1BefRmBMc" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen&gt;&lt;/iframe&gt;

 [Eventlow](https://hcil.umd.edu/eventflow/)
    </textarea>
 <style data-target="print-only">@media screen {.remark-slide-container{display:block;}.remark-slide-scaler{box-shadow:none;}}</style>
 <script src="https://remarkjs.com/downloads/remark-latest.min.js"></script>
 <script src="addons/macros.js"></script>
 <script>var slideshow = remark.create({
@@ -177,57 +195,16 @@ class: full, center
 if (window.HTMLWidgets) slideshow.on('afterShowSlide', function (slide) {
  window.dispatchEvent(new Event('resize'));
 });
 (function(d) {
  var s = d.createElement("style"), r = d.querySelector(".remark-slide-scaler");
 (function() {
  var d = document, s = d.createElement("style"), r = d.querySelector(".remark-slide-scaler");
  if (!r) return;
  s.type = "text/css"; s.innerHTML = "@page {size: " + r.style.width + " " + r.style.height +"; }";
  d.head.appendChild(s);
 })(document);

 (function(d) {
  var el = d.getElementsByClassName("remark-slides-area");
  if (!el) return;
  var slide, slides = slideshow.getSlides(), els = el[0].children;
  for (var i = 1; i < slides.length; i++) {
    slide = slides[i];
    if (slide.properties.continued === "true" || slide.properties.count === "false") {
      els[i - 1].className += ' has-continuation';
    }
  }
  var s = d.createElement("style");
  s.type = "text/css"; s.innerHTML = "@media print { .has-continuation { display: none; } }";
  d.head.appendChild(s);
 })(document);
 // delete the temporary CSS (for displaying all slides initially) when the user
 // starts to view slides
 (function() {
  var deleted = false;
  slideshow.on('beforeShowSlide', function(slide) {
    if (deleted) return;
    var sheets = document.styleSheets, node;
    for (var i = 0; i < sheets.length; i++) {
      node = sheets[i].ownerNode;
      if (node.dataset["target"] !== "print-only") continue;
      node.parentNode.removeChild(node);
    }
    deleted = true;
  });
 })();</script>

 <script>
 (function() {
  var links = document.getElementsByTagName('a');
  for (var i = 0; i < links.length; i++) {
    if (/^(https?:)?\/\//.test(links[i].getAttribute('href'))) {
      links[i].target = '_blank';
    }
  }
 })();
 </script>

 <script>
 slideshow._releaseMath = function(el) {
  var i, text, code, codes = el.getElementsByTagName('code');
  var i, text, code, codes = document.getElementsByTagName('code');
  for (i = 0; i < codes.length;) {
    code = codes[i];
    if (code.parentNode.tagName !== 'PRE' && code.childElementCount === 0) {
@@ -241,8 +218,7 @@ slideshow._releaseMath = function(el) {
    }
    i++;
  }
 };
 slideshow._releaseMath(document);
 })();
 </script>
 <!-- dynamically load mathjax for compatibility with self-contained -->
 <script>
--- a/courses/bar-chart-race.Rmd
+++ b/courses/bar-chart-race.Rmd
@@ -26,13 +26,12 @@ library(gganimate)
 Load data
 ```{r load_data}
 # load data
 data("base_pop")
 data("us_city_populations")

 # create the ranks based on census populations
 base_pop<- base_pop %>%
  group_by( annee) %>%
  arrange(desc(population)) %>%
  mutate(rang = 1:n()) %>%
 # create the ranks based on census Populations
 us_city_populations<-us_city_populations %>%
  group_by( Year) %>%
  arRanke(desc(Population)) %>%
  ungroup()

 ```
@@ -42,8 +41,8 @@ Extract the list of top n cities
 ```{r extract_top_n}
 n_cities = 10

 top_cities <- base_pop %>% filter(rang <= n_cities) %>%
  select(ville, dep, region) %>% distinct()
 top_cities <-us_city_populations %>% filter(Rank <= n_cities) %>%
  select(City, State, Region) %>% distinct()

 ```

@@ -51,31 +50,31 @@ top_cities <- base_pop %>% filter(rang <= n_cities) %>%
 Create all missing dates
 ```{r, combine_dates}
 # create a data frame with all the years between min and max Year
 all_years <- data.frame(annee = seq(min(base_pop$annee), max(base_pop$annee), 1))
 all_years <- data.frame(Year = seq(min(us_city_populations$Year), max(us_city_populations$Year), 1))

 # combine top_cities and all_years
 all_combos <- merge(top_cities, all_years, all = T)

 # combine all_combos with the original dataset
 res_interp <- merge(base_pop, all_combos, all.y = T)
 res_interp <- merge(us_city_populations, all_combos, all.y = T)
 ```

 Interpolate the populations when missing (linear interpolation here)
 Interpolate the Populations when missing (linear interpolation here)
 ```{r, interpolate}
 res_interp  <- res_interp  %>%
  group_by(ville) %>%
  mutate(population=approx(annee,population,annee)$y)
  group_by(City) %>%
  mutate(Population=approx(Year,Population,Year)$y)
 ```


 ```{r, filter_for_plot}
 # filter the top ten cities per year
 to_plot <- res_interp %>%
  group_by(annee) %>%
  arrange(-population) %>%
  mutate(rang=row_number()) %>%
  filter(rang<=n_cities) %>%
  rename(`Région` = region)
  group_by(Year) %>%
  arrange(-Population) %>%
  mutate(Rank=row_number()) %>%
  filter(Rank<=n_cities) %>%
  rename(`Région` = Region)

 ```

@@ -84,10 +83,10 @@ Make the plot
 ```{r, make_plot}
 p <- to_plot %>%

  ggplot(aes(x = -rang,y = population, group =ville)) +
  geom_tile(aes(y = population / 2, height = population, fill = `Région`), width = 0.9) +
  geom_text(aes(label = ville), hjust = "right", colour = "white", fontface="bold", nudge_y = -10000) +
  geom_text(aes(label = scales::comma(population,big.mark = ' ')), hjust = "left", nudge_y = 10000, colour = "grey90") +
  ggplot(aes(x = -Rank,y = Population, group =City)) +
  geom_tile(aes(y = Population / 2, height = Population, fill = `Région`), width = 0.9) +
  geom_text(aes(label = City), hjust = "right", colour = "white", fontface="bold", nudge_y = -100000) +
  geom_text(aes(label = scales::comma(Population,big.mark = ' ')), hjust = "left", nudge_y = 100000, colour = "grey90") +
  coord_flip(clip="off") +
  hrbrthemes::scale_fill_ipsum() +
  scale_x_discrete("") +
@@ -104,17 +103,16 @@ p <- to_plot %>%
    legend.text = element_text(size = 15),
    legend.background = element_blank()) +
  # gganimate code to transition by year:
  transition_time(annee) +
  transition_time(Year) +
  ease_aes('cubic-in-out') +
  labs(title='Evolution des plus grandes villes de France',
       subtitle='Population en {round(frame_time,0)}',
       caption='Source: INSEE Base populations historiques 1876-2015')
  labs(title='Evolution des plus grandes villes US',
       subtitle='Population en {round(frame_time,0)}')

 ```

 ```{r, animate}
 animate(p, nframes = 350, fps = 25, end_pause = 30, width = 1200, height = 1200, start_pause = 15 )

 animate(p, nframes = 300, fps = 25, end_pause = 30, width = 1200, height = 1200, start_pause = 15 ,)
 anim_save("bar_race.gif", animation = last_animation())
 ```


--- a/courses/bar_race.gif
+++ b/courses/bar_race.gif
--- a/courses/lab01-correction.Rmd
+++ b/courses/lab01-correction.Rmd
@@ -0,0 +1,106 @@
 ---
 title: "correction"
 author: "Antoine Neuraz"
 date: "18/11/2019"
 output: html_document
 ---

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

 ## Ouvrir le dataset "mtcars"

 ## représenter le "Gross horsepower" en fonction du nombre de cylindres

 ```{r}
 data("mtcars")

 ggplot(data = mtcars, 
       aes(x = as.factor(cyl),
           y = hp)) + 
  geom_jitter(width = .2)
 ```


 ## utiliser l'encodage multiple sur le nombre de cylindres

 ```{r}
 ggplot(data = mtcars, 
       aes(x = as.factor(cyl),
           y = hp, 
           size = cyl,
           color = cyl)) + 
  geom_jitter(width = .2, alpha = .6) +
  theme_minimal() +
  theme(legend.position = "none")
 ```

 ## ajouter l'information du nombre de carburateurs

 ```{r}

 ggplot(data = mtcars, 
       aes(x = as.factor(cyl),
           y = carb, 
           size = hp,
           color = hp)) + 
  geom_jitter(width = .2, alpha = .6) +
  theme_minimal() 
  #facet_grid(~as.factor(carb))
  #theme(legend.position = "none")
 ```

 ## Paufiner le plot (axes, titres, thème)

 ```{r}

 ggplot(data = mtcars, 
       aes(x = as.factor(cyl),
           y = carb, 
           size = hp,
           color = hp)) + 
  geom_jitter(width = .2, alpha = .6) +
  theme_minimal() + 
  labs(x = "Cylinders", 
       y = "Carburators")
 ```

 ## représenter la distribution du nombre de miles per gallon en histogramme

 ```{r}
 ggplot(mtcars, 
       aes(x= mpg)) + 
  geom_histogram(bins = sqrt(nrow(mtcars)))
 ```

 ## représenter la distribution du nombre de miles per gallon en boxplot

 ```{r}
 ggplot(mtcars, 
       aes(x= 1, y= mpg)) + 
  geom_boxplot()
 ```

 ## representer la distribution du nombre de miles per gallon en fonction du nombre de cylindres

 ```{r}
 ggplot(mtcars, 
       aes(x= as.factor(cyl), y= mpg)) + 
  geom_violin(fill = "grey70")
 ```

 ## ajouter les points par dessus la distribution

 ```{r}
 ggplot(mtcars, 
       aes(x= as.factor(cyl), y= mpg)) + 
  geom_violin(fill = "grey70") + 
  geom_jitter(aes(color = cyl),width = .15) 
 ```


 ## paufiner le plot (axes, titres, thème)



--- a/courses/lab7-temporal_data.Rmd
+++ b/courses/lab7-temporal_data.Rmd
@@ -0,0 +1,127 @@
 ---
 title: "Lab 07 - Données temporelles et textuelles"
 author: "Antoine Neuraz"
 date: "22/11/2019"
 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 = .7, fig.width = 12)
 library(vizoR)
 library(ggplot2)
 library(gghighlight)
 library(dplyr)
 library(ggTimeSeries)
 ```

 ## TODO

 #### 1. charger le dataset `us_city_populations` de la librairie `vizoR`

 #### 2. tracer un line chart de l'évolution de la population des villes US

 #### 3. Mettez en évidence les 5 plus grandes villes (hint: package gghighlight)
 [introduction gghighlight](https://cran.r-project.org/web/packages/gghighlight/vignettes/gghighlight.html)

 #### 4. Appliquez les principes de design de Tufte 

 ##### 5. BONUS: affichez le nom des villes directement à la fin de la ligne

 #### 6. Réalisez un streamgraph des 5 plus grandes villes US (hint: package ggTimeSeries)

 ---

 ## TODO 2

 #### Trouver une 3e visualization pertinente pour montrer l'évolution de la population des villes US. 

 ---

 ```{r}

 data("us_city_populations")

 n_cities = 5

 # top_cities <- 
 #   us_city_populations %>% 
 #   filter(Rank <= n_cities) %>%
 #   select(City, State, Region) %>% 
 #   distinct()
 # 
 # to_plot <- filter(us_city_populations, City %in% top_cities$City)

 #to_plot <- us_city_populations

 last_ranks <- us_city_populations %>% 
  filter(Year == max(Year)) %>%
  mutate(last_rank = Rank) %>% 
  select(City, last_rank)

 to_plot <- left_join(us_city_populations, last_ranks, by= 'City')

 right_axis <-  to_plot %>% 
  group_by(City) %>% 
  top_n(1, Year) %>% 
  ungroup() %>%
  top_n(n_cities, -last_rank)

 ends <- right_axis %>%
  pull(Population) 

 labels <-  right_axis %>%
  pull(City)

 ```

 ---
 class: full
 ```{r, echo = FALSE}
 ggplot(to_plot, aes(x=Year, y = Population, group = City, color = City)) + 
  geom_line(size=1) + 
  #geom_text(data = subset(to_plot, Year == 2010), aes(x=Inf, y = Population, label=City), hjust = 1) +
  scale_x_continuous("", expand=c(0,0))+
  scale_y_continuous("",
                     labels=scales::comma_format(big.mark = " "),
                     sec.axis = sec_axis(~ ., breaks = ends, labels = labels ))+
  scale_color_viridis_d()+
  theme_elegant_dark()+
  theme(legend.position = "none",
        plot.margin = unit(c(1,3,1,1), "lines"), 
        axis.line.y = element_blank(),
        axis.line.x = element_blank(), 
        axis.ticks.x = element_line(), 
        panel.grid.major.y = element_line(color= 'grey30', size = .2) ) + 
  gghighlight(max(last_rank) <= n_cities, use_direct_label = FALSE, label_key = City,unhighlighted_colour = "grey20")
  
 ```
 ---
 class: full
 ```{r, echo = FALSE}
 library(ggTimeSeries)
 to_plot %>% filter(City %in% labels) %>%
  ggplot(aes(x = Year, y = Population, group = City, fill = City)) +
  scale_y_continuous("", labels = scales::comma_format(big.mark = " "))+
  stat_steamgraph() +
  theme_elegant_dark() + 
  scale_fill_viridis_d() + 
  theme(plot.margin = unit(c(1,3,1,1), "lines"), 
        axis.line.y = element_blank(),
        axis.line.x = element_blank(), 
        axis.ticks.x = element_line(), 
        panel.grid.major.y = element_line(color= 'grey30', size = .2) ) 

 ```
 ---

 ![](bar_race.gif)

--- a/courses/lab7-temporal_data.html
+++ b/courses/lab7-temporal_data.html
@@ -0,0 +1,132 @@
 <!DOCTYPE html>
 <html>
  <head>
    <title>Lab 07 - Données temporelles et textuelles</title>
    <meta charset="utf-8">
    <meta name="author" content="Antoine Neuraz" />
    <link href="libs/remark-css-0.0.1/default.css" rel="stylesheet" />
    <link rel="stylesheet" href="css/my_style.css" type="text/css" />
  </head>
  <body>
    <textarea id="source">




 ## TODO

 #### 1. charger le dataset `us_city_populations` de la librairie `vizoR`

 #### 2. tracer un line chart de l'évolution de la population des villes US

 #### 3. Mettez en évidence les 5 plus grandes villes (hint: package gghighlight)
 [introduction gghighlight](https://cran.r-project.org/web/packages/gghighlight/vignettes/gghighlight.html)

 #### 4. Appliquez les principes de design de Tufte 

 ##### 5. BONUS: affichez le nom des villes directement à la fin de la ligne

 #### 6. Réalisez un streamgraph des 5 plus grandes villes US (hint: package ggTimeSeries)

 ---

 ## TODO 2

 #### Trouver une 3e visualization pertinente pour montrer l'évolution de la population des villes US. 

 ---


 ```r
 data("us_city_populations")

 n_cities = 5

 # top_cities &lt;- 
 #   us_city_populations %&gt;% 
 #   filter(Rank &lt;= n_cities) %&gt;%
 #   select(City, State, Region) %&gt;% 
 #   distinct()
 # 
 # to_plot &lt;- filter(us_city_populations, City %in% top_cities$City)

 #to_plot &lt;- us_city_populations

 last_ranks &lt;- us_city_populations %&gt;% 
  filter(Year == max(Year)) %&gt;%
  mutate(last_rank = Rank) %&gt;% 
  select(City, last_rank)

 to_plot &lt;- left_join(us_city_populations, last_ranks, by= 'City')

 right_axis &lt;-  to_plot %&gt;% 
  group_by(City) %&gt;% 
  top_n(1, Year) %&gt;% 
  ungroup() %&gt;%
  top_n(n_cities, -last_rank)

 ends &lt;- right_axis %&gt;%
  pull(Population) 

 labels &lt;-  right_axis %&gt;%
  pull(City)
 ```

 ---
 class: full
 ![](lab7-temporal_data_files/figure-html/unnamed-chunk-2-1.png)&lt;!-- --&gt;
 ---
 class: full
 ![](lab7-temporal_data_files/figure-html/unnamed-chunk-3-1.png)&lt;!-- --&gt;
 ---

 ![](bar_race.gif)
    </textarea>
 <script src="https://remarkjs.com/downloads/remark-latest.min.js"></script>
 <script src="addons/macros.js"></script>
 <script>var slideshow = remark.create({
 "ratio": "4:3",
 "countIncrementalSlides": false,
 "self-contained": true
 });
 if (window.HTMLWidgets) slideshow.on('afterShowSlide', function (slide) {
  window.dispatchEvent(new Event('resize'));
 });
 (function() {
  var d = document, s = d.createElement("style"), r = d.querySelector(".remark-slide-scaler");
  if (!r) return;
  s.type = "text/css"; s.innerHTML = "@page {size: " + r.style.width + " " + r.style.height +"; }";
  d.head.appendChild(s);
 })();</script>

 <script>
 (function() {
  var i, text, code, codes = document.getElementsByTagName('code');
  for (i = 0; i < codes.length;) {
    code = codes[i];
    if (code.parentNode.tagName !== 'PRE' && code.childElementCount === 0) {
      text = code.textContent;
      if (/^\\\((.|\s)+\\\)$/.test(text) || /^\\\[(.|\s)+\\\]$/.test(text) ||
          /^\$\$(.|\s)+\$\$$/.test(text) ||
          /^\\begin\{([^}]+)\}(.|\s)+\\end\{[^}]+\}$/.test(text)) {
        code.outerHTML = code.innerHTML;  // remove <code></code>
        continue;
      }
    }
    i++;
  }
 })();
 </script>
 <!-- dynamically load mathjax for compatibility with self-contained -->
 <script>
 (function () {
  var script = document.createElement('script');
  script.type = 'text/javascript';
  script.src  = 'https://mathjax.rstudio.com/latest/MathJax.js?config=TeX-MML-AM_CHTML';
  if (location.protocol !== 'file:' && /^https?:/.test(script.src))
    script.src  = script.src.replace(/^https?:/, '');
  document.getElementsByTagName('head')[0].appendChild(script);
 })();
 </script>
  </body>
 </html>
--- a/courses/lab7-temporal_data_files/figure-html/unnamed-chunk-2-1.png
+++ b/courses/lab7-temporal_data_files/figure-html/unnamed-chunk-2-1.png
--- a/courses/lab7-temporal_data_files/figure-html/unnamed-chunk-3-1.png
+++ b/courses/lab7-temporal_data_files/figure-html/unnamed-chunk-3-1.png
--- a/courses/line-charts.Rmd
+++ b/courses/line-charts.Rmd
@@ -1,124 +0,0 @@
 ---
 title: "Line charts for population data"
 output: rmarkdown::html_vignette
 vignette: >
  %\VignetteIndexEntry{bar-chart-race}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
 ---

 ```{r setup, include=FALSE}
 knitr::opts_chunk$set(echo = TRUE)
 library(vizoR)
 library(ggplot2)
 library(gghighlight)
 library(dplyr)
 ```


 ```{r}

 data("us_city_populations")

 n_cities = 5

 top_cities <- us_city_populations %>% filter(Rank <= n_cities) %>%
  select(City, State, Region) %>% distinct()

 to_plot <- filter(us_city_populations, City %in% top_cities$City)
 to_plot <- us_city_populations

 last_ranks <- to_plot %>% filter(Year == max(Year)) %>%
  mutate(last_rank = Rank) %>% select(City, last_rank)

 to_plot <- left_join(to_plot, last_ranks, by= 'City')

 right_axis <-  to_plot %>% 
  group_by(City) %>% 
  top_n(1, Year) %>% 
  ungroup() %>%
  top_n(n_cities, -last_rank)

 ends <- right_axis %>%
  pull(Population) 

 labels <-  right_axis %>%
  pull(City)

 ```

 ```{r}
 ggplot(to_plot, aes(x=Year, y = Population, group = City, color = Region)) + 
  geom_line(size=1) + 
  #geom_text(data = subset(to_plot, Year == 2010), aes(x=Inf, y = Population, label=City), hjust = 1) +
  scale_x_continuous("", expand=c(0,0))+
  scale_y_continuous("",
                     labels=scales::comma_format(big.mark = " "),
                     sec.axis = sec_axis(~ ., breaks = ends, labels = labels ))+
  theme_elegant() + 
  theme(legend.position = "bottom",
        plot.margin = unit(c(1,3,1,1), "lines"), 
        axis.line.y = element_blank(),
        axis.line.x = element_blank(), 
        axis.ticks.x = element_line(), 
        panel.grid.major.y = element_line(color= 'grey90', size = .2) ) + 
  gghighlight(max(last_rank) <= n_cities, use_direct_label = FALSE, label_key = City)
  
 ```

 ```{r}

 data("base_pop")

 n_cities = 10

 # create the ranks based on census populations
 base_pop<- base_pop %>%
  group_by( annee) %>%
  arrange(desc(population)) %>%
  mutate(rang = 1:n()) %>%
  ungroup()

 top_cities <- base_pop %>% filter(rang <= n_cities) %>%
  select(ville, dep , region) %>% distinct()

 to_plot <- filter(base_pop, ville %in% top_cities$ville)
 #to_plot <- base_pop

 last_ranks <- to_plot %>% filter(annee == max(annee)) %>%
  mutate(last_rank = rang) %>% select(ville, last_rank)

 to_plot <- left_join(to_plot, last_ranks, by= 'ville')

 right_axis <-  to_plot %>% 
  group_by(ville) %>% 
  top_n(1, annee) %>% 
  ungroup() %>%
  top_n(n_cities, -last_rank)

 ends <- right_axis %>%
  pull(population) 

 labels <-  right_axis %>%
  pull(ville)

 ```

 ```{r}
 ggplot(to_plot, aes(x=annee, y = population, group = ville, color = region)) + 
  geom_line(size=1) + 
  #geom_text(data = subset(to_plot, Year == 2010), aes(x=Inf, y = Population, label=City), hjust = 1) +
  scale_x_continuous("", expand=c(0,0))+
  scale_y_continuous("",
                     labels=scales::comma_format(big.mark = " "),
                     sec.axis = sec_axis(~ ., breaks = ends, labels = labels ))+
  theme_elegant() + 
  theme(legend.position = "bottom",
        plot.margin = unit(c(1,3,1,1), "lines"), 
        axis.line.y = element_blank(),
        axis.line.x = element_blank(), 
        axis.ticks.x = element_line(), 
        panel.grid.major.y = element_line(color= 'grey90', size = .2) ) + 
  gghighlight(max(last_rank) <= n_cities, use_direct_label = FALSE, label_key = ville)
  
 ```